Brad Geddes / PPC Geek
Official AdWords Seminar Leader.
Author of Advanced Google AdWords.
Co-Founder, Adalysis.
(312) 884-9017

Blog.

Should You Use Exact Match Keywords?

I recently saw a discussion on how match types are dead and how you should only use modified broad match. As exact match shows for variants, you don’t really control your search queries; therefore, only use modified broad match. At least that was the argument.

Initially I dismissed this discussion until I saw one of the comments state that managing match types is old school and that it is not the new reality – the new reality is that only modified broad match matters. Having been in this industry for 17 years, I thought maybe I was living in my old school ways and I just needed to check into some data.

Account 1: Goes Beyond Match Type Control – Singular & Plural Control

So I started by analyzing an account that I knew was an extreme outlier; but I thought would prove out my thoughts on match types and why they matter as well as their underlying search queries.

I don’t want to reveal the identity of this advertiser, so I’m going to ‘widgetize’ the data. However, to give you some ideas about the company – they like to sell in bulk. If you need 500 uniforms or 1000 clogs, they can help you out. While they will sell you one pair – that’s not their goal. Early on, we saw some data really showing how plurals and singulars have different behavior for them and not only manage the match types by campaigns, the singulars and plurals are managed by ad group.

For instance, their exact match ad groups are set up like:

Ad Group Keywords – Exact Negative Keywords
Blue widgets (p) Blue widgets one
Blue widgets two
Blue widgets three
widget
Blue Widget (s) Blue widget one
Blue widget two
Blue widget three
widgets

As you can see, ad group (p) only shows for the plural versions and ad group (s) only shows for the singular versions.

Now, when we look at the conversion rates and average order values; we see why it’s organized into singulars and plurals:

Ad Group Conv Rate Average Order Value
Blue widgets (p) 2.3% $673
Blue Widget (s) 2.9% $81

 

They are willing to bid six times higher for the plural versions than the singular versions due to the huge difference in average order value.

If they just put all these keywords into the same ad group and used keyword level bids; the organization would not work since the singular and plural queries could show for each keyword due to variation or near match (a search query can be a singular, plural, or misspelling even for your exact match terms).

In their case, they have more than 500 products where the average order value changes by singular and plural terms (meaning there are 500 singulars and 500 plural ad groups). In fact, this singular and plural version is something we often see with sports teams. Do you buy a Seattle Seahawks uniform or a Seattle Seahawk uniform – what does that little ‘s’ tell you about the searcher? Are they a fan or someone buying a gift who knows nothing about sports teams?

At this point in time – we haven’t actually addressed the issue if they should manage the exact match separately from the broad match. We just know that there is an average order value difference.

If this company spends $100/month on PPC. Managing these independently probably won’t matter as the extra money and time spent on management fees would outstrip the complexity gains. If this company spends $50,000/month on PPC and has PPC revenues over $250,000/month; then it is worth the extra fees to manage these queries independently.

If they were limited by budget, they want 100% impression share on the plurals and the rest over budget can be spent on the singulars.

Account 2: Throw It All Together and See What Comes Out

There are times when managing exact match and the query data doesn’t matter as much. This account spends about $10,000/month; and outside of broad match – the data is pretty similar by match types:

Match Type Conv Rate Average Order Value
Exact 9.3% $423
Phrase 9.1% $412
Modified Broad 9.2% $419
Broad 6.23% $441

 

If you start segmenting the account by singulars and plurals; there is no difference in the data. The only way to really see any type of a difference in queries is if start segmenting the data by individual query by product sold (please note it is a mature account and is setup incredibly well). If this account managed all the queries perfectly, it would take another 15-20 hours/month and it would make another $2,000/month.

That does seem like a huge time difference for just managing some queries; and if you wonder why, it’s due to the current structure. This account has four main search campaigns.

  • One campaign uses CPA bidding where the bids are set by ad group based upon the sales by ad group and includes exact and modified broad matches.
    • The geo area is their ‘top performing regions’
  • The second and third campaigns are identical to the first campaign, except the ads (through testing) are slightly different, and the ads are shown in regions that don’t do as well (but still perform well) as their main regions.
  • The last campaign has a much lower budget, utilizes broad match, some DSAs, and is considered a backfill campaign.

Because the account uses CPA bidding (and it works very well for this account once the account was geographically segmented) if this account moved more towards perfecting its query serving and match type usage; it would have to move to a different bid system (which is more time or money depending on how it’s done).

If this account was spending $100,000 or more per month so that managing the queries becomes an additional $20,000 in revenue – then suddenly managing the queries and match types makes a lot of sense and is something this account would do.

So – Should You Use Multiple Match Types & Control Their Usage?

It’s not about match types – it is about return on time and return on management dollars when you get into complex organizations.

In the first case, the company makes a lot more money by paying someone to manage a very complex structure that manages not just match types, but singular and plural versions due to how their account functions.

In the second case, the company makes more money by not spending all their time managing their account, and by spending that time perfecting their customer service.

In the end – ALL absolute statements about match types being dead — are wrong.

Match types live on.

The management fees and time to manage the match types is where the discussion should be and if it is worth the time and money to get that granular in your accounts.

The Complete AdWords Audit Part 13: Bid Management

This is a continuation of the AdWords Audit Series. You can see previous parts here: Introduction, Goal setting, Measurement, Campaign Settings & Bid Adjustments, Ad Extensions, Impression Share & Auction Insights, Quality Score, Account Structure, Keywords & Match Types, Ad Copy, Testingthe Lin-Rodnitzky Ratio and Google Shopping Campaigns.

 

“Great bids can’t fix a lousy campaign”

Craig Danuloff

Whenever I think of bid management, I think of this quote from Craig Danuloff, back in 2010 in his excellent e-book ‘The 21 Secret Truths of High Resolution PPC’.

In the original e-book, Craig wrote the following about bid management:

“Bidding gets entirely too much attention in the paid search world. Both the idea of it and its process are treated as if they’re the central element to successful campaigns.
The truth is that bidding is just another piece of the puzzle. Even the best bid strategy or most powerful bid algorithm can’t overcome poor keyword selection, sloppy organization, or uninviting ad copy.

Think of it this way: there are pre-requisites for bidding. Until a keyword is ready, anything more than a simple target-position bid strategy is most likely a waste of time and perhaps money.

Don’t worry too much about the bid for any keyword until it:

  • Lives in a tightly organized ad group
  • Uses the right match type
  • Attracts appropriate search queries
  • Is paired with targeted and persuasive text ad copy
  • Sends traffic to an effective landing page or conversion path

Any bids calculated and set before these milestones are achieved are based on inconsistent inputs, so they really can’t be accurate or optimal. Watch carefully and you’ll probably notice prematurely applied algorithmic bids that are too high, too low, or change radically from one period to another.

Garbage in, garbage out. As they say.

On the other hand, when the relationship between search queries, keywords, and text ads has been thoughtfully managed or optimized, smart bid strategies or advanced algorithms can be an incredibly effective way to maximize sales and revenues.

Spend a lot of time on bidding… after you’ve spent a lot of time on everything else.”

I believe all of the above is as true today as it was in 2010. Although I’d rather not suggest to spend a lot of time on bidding. Ideally, try to spend no more than 20% of your time and resources on bidding, and preferably a bit less.
I’m not saying bidding isn’t important, because it is. Your bids have a large influence on your position and CPC (as has Quality Score), which on their turn have a large influence on your CTR and ROI. So those are some good reasons to care about your bids.

On the other hand, there are 2 reasons why you don’t want to spend too much time on calculating and changing bids:

  • Calculating bids is basically number-crunching, and as we know, with enough (predictable) data, software can do that better and faster than humans can.
  • A new bid doesn’t give you a long-term advantage over your competition as they can easily change their bids as well.

What does give you a longer term and even unfair advantage is having higher Quality Scores and/or conversion rates than your competition. In that case you’ll be paying less for the same click while getting more in return.
So it’s best to avoid tinkering with your bids all day and focus more on creative and strategic work that can’t be automated instead. Especially if that work can lead to the unfair advantages of higher Quality Scores and conversion rates, like writing great ad copy and improving your landing pages.

In this series, I’ve tried to provide guidelines that are sensible for most AdWords accounts in most cases, so that readers can check their own accounts and implement whatever advice makes sense for them.

However, when it comes to answering the question: what’s the best way to manage my bids? The answer really is: it depends. So I can’t offer a one size fits all solution today, but the goal still is to provide best practices that are sensible for your specific situation.

That’s why the rest of this post is divided in the following parts:

  1. Before you start: profit-driven goals, optimal tracking and attribution.
  2. Setting bids for new keywords.
  3. Manual bidding with formula’s, filters, rules, scripts and the bid simulator. These methods are probably best for lower volume accounts or campaigns, generating less than 60 to 100 non-branded conversions a month, spending less than $10,000 a month or having no more than a few hundred keywords that matter.
    It’s also the way to go if you want to keep full control of your bids, especially during sales or events like Black Friday, Cyber Monday, Super Bowl, Valentine’s Day, etc.
  4. Automated bidding with AdWords flexible bid strategies. This method is probably best for medium to high volume accounts or campaigns, generating more than 60 to 100 non-branded conversions a month. The more conversions you have, the better it works.
  5. Automated bidding with 3rd party technology. This method is probably best for high volume enterprise accounts, generating at least 500 non-branded conversions a month, spending $50,000 – $100,000 a month or more or having thousands of keywords that matter. It especially makes sense if you also advertise significantly outside of AdWords, e.g. on Bing, Yahoo!, Facebook or Display Networks outside the GDN.

This will make it the longest post in this series by far, which is ironic after downplaying the importance of bid management in the introduction. Nonetheless, it’s a complex subject with a lot of possibilities, so I hope you’ll bear with me and will find useful tips to apply to your campaigns and bids!

 

  1. Before you start: profit-driven goals, optimal tracking and attribution

Before you can even calculate bids, whether it’s manual or automated, you need to have the following in place:

  • A profitable efficiency target like CPA or ROAS
  • Tracking all conversions and/or value generated by your campaigns

You could have multiple CPA or ROAS targets in your account (e.g. because of different margins) and that’s perfectly fine.
You could also care more about clicks, position or visibility than about conversions or revenue, and that’s also fine (well… sort of).
However, this post will not go any deeper into awareness-based bidding besides mentioning the automated flexible bid strategies that can help you with that: maximize clicks, target search page location and target outranking share.

If you choose for any of these awareness-based strategies, please realize that conversion rates don’t vary (much) by position, and if they do, it’s more in favor of the 2nd and 3rd position (and not of the 1st position).
But it’s generally best to just bid based on value per click, improve your Quality Scores as much as you can and let the positions fall where they may.

Profit-driven goals

It’s extremely important that your non-branded CPA or ROAS target takes the following into account:

  • Your margin
  • The part of this margin you’re willing to invest in acquisition

Based on these 2 inputs, you can easily calculate your target CPA or ROAS, as explained in the Goal Setting part of this series.

And to be truly profit-driven, you need to acknowledge the trade-off between volume and efficiency. This means you’re trying to find the target CPA or ROAS that maximizes your total profit in real Dollars, Euros, Pounds or whatever currency pays your bills.
Which isn’t the same as minimizing your CPA or maximizing your ROAS or ROI. Or as Google explains it in their Profit-Driven Marketing articles:

“Would you prefer an $80 CPA or a $90 CPA?

It’s actually a trick question, and the profit-driven marketer would say, “I don’t have enough information to choose between them.” Perhaps the $90 CPA lands your ad in the top spot on a search results page and brings way more sales volume than the lower-priced CPA unit. In that case, the correct decision is to spend more to make more.”

cpa-profit-volume-vs-efficiency

You can see the visualization of this example in the graph above. Obviously, if your current CPA target was $100 in this example, then the profit-driven thing to do was to lower it to $90. Somehow, Google forgot to mention that ;-).

The only way to find out your profit-maximizing target is to test with different CPA or ROAS targets (at least 20% to 30% higher or lower than your current target to make sure there’s a significant impact) and find out which combination of CPA or ROAS and conversions or revenue delivers you the most profit.

Optimal Tracking

As mentioned in the Measurement part of this series, you should track the full value that your campaigns deliver, so all this value can be taken into account when setting bids:

  • Obviously your macro conversions, typically leads or the revenue from sales
  • Cross-device conversions, which usually means an uplift of 5% to 15% in conversions and can be integrated with flexible bid strategies
  • The lifetime value of your customers
  • The quality of your leads (if you’re lead-gen)
  • Micro conversions, such as newsletter subscriptions

And if applicable to your business, you should also track the value of:

  • Phone calls
  • App downloads
  • In-store visits

So if your goals and/or tracking aren’t optimal, it’s impossible for your bids to be. Being the 2 main inputs for any bid management, your goal setting and tracking need to be as accurate and complete as they can possibly be.

Attribution

And then of course, there’s attribution. Most of us are still bidding based on last-click attribution, and that’s mainly because it’s hard to:

  • Agree on an attribution model that isn’t last click.
  • Manually adjust your bids and budgets based on this attribution model .

The very least you should do, is ignore the last branded click if this was preceded by a non-branded click, a so called ‘last touch with brains’ model. If you don’t do this and have a lot of branded conversions in your account, you’re probably undervaluing your non-branded keywords and campaigns.

Once you do that, it’s not a complete disaster if the rest of your keywords are judged mostly on their last-click performance.
As Siddharth Shah wrote a few years ago in his analysis ‘Attribution: Busting the Myths’: “In short, if you are doing search primarily, changing attribution rules usually doesn’t change much. You typically see a 10-20% assist funnels and 5-10% non-brand to brand type funnels. For the most part, you are fine working with the last click. However, when looking at multi-channel data, when search forms a small component of the overall picture, multi-event attribution can matter.”

This is also what other experts found when digging into to the attribution reports (formerly search funnels): only a small percentage of keywords can justify their existence based purely on assists. In other words: attribution is no excuse to keep ‘upper funnel keywords’ running that are highly inefficient. If you want awareness, effective display campaigns will offer you just that, at CPM’s that are hundreds of times lower.
Or as George Michie states in one of his comments in the Cardinal Sin of Paid Search: paid search is a very expensive channel for brand building. If we think of an advertising impression being a visit to your website (I don’t think there’s much of an impression given by the little text ads), then a $0.50 click translates to a $500 CPM. That’s an awfully expensive way to show people you exist.”

And for those of us that would like to be more sophisticated when it comes to attribution, but don’t want to lose a lot of time recalculating bids, there’s good news.
In their livestream in May this year, Google announced that advertisers will be able to select an attribution model (including data-driven attribution) in AdWords and that automated bidding will then be based on the chosen model.
Watch the part from my fellow Dutchman Joan Arensman about making attribution actionable to see how this would work.

adwords-data-driven-attribution

Let’s hope this will be released soon, because until then, the alternative is to do quite some manual work analyzing the attribution reports and trying to calculate bids based on your chosen attribution model.

  1. Setting bids for new keywords

If you just added a new keyword to your AdWords account, you may wonder what would be a good starting bid. As you have no click or conversion data yet, there are several options in this case:

  • The estimates you get for that keyword from Google’s Keyword Planner
  • Somewhere between the first page bid and the top of page bid estimate. But be careful: these estimates are often (way) too high when you just added a keyword. So wait at least a couple of hours (or days) before using these estimates.
  • The expected value per click based on your expected conversion rate or revenue per click for that keyword. You’ll find the formulas to calculate the value per click below.

For more details about this topic, be sure to read Brad Geddes’ 4 Ways to Determine Your Starting Bids.

Depending on your budget, aim for an average position between 2 and 4 to start with, so you can accumulate the necessary data to refine your bids later on.

  1. Manual bidding with formula’s, filters, rules, scripts and the bid simulator

As stated in the beginning, manual bidding is best for lower volume accounts. For the purpose of this post, I’ll define lower volume as having less than 60 to 100 non-branded conversions a month. You’ll see why in the part about flexible bid strategies.
Manual bidding is fine for lower volume accounts because the time you need to calculate the right bids is limited and because you probably won’t benefit from the advantages of automated bidding, as automation works best with lots of (predictable) data.

Manual bidding can also be a (temporary) necessity during major sales, as algorithms will need at least a day (or more) to adjust to the effects of sudden spikes in volume, competition or conversion rates. That could mean the sale is already over once the bids have been adjusted to the right levels.

Value per click formula’s

Once a keyword has converted, you can easily calculate its value per click based on your target CPA or ROAS. Obviously, you’d rather have multiple conversions so your bids are based on solid data. Ideally, you’d calculate the bid based on the last 10 to 20 conversions.
For some keywords that means looking back 7 days and for other keywords you may need to look back 12 months.

The formulas to calculate the value per click of a keyword are as following:

For CPA targets:
(keyword conversion rate) x Target CPA

So if a keyword converts at 1.8% and your target CPA is $75, then the value per click of that keyword is 1.8% x $75 = $1.35.
This is the CPC you’re willing to pay to reach your target CPA with this keyword. You could bid a bit higher than this as your average CPC will be lower than your max CPC.

For ROAS targets:
(average revenue per click) / Target ROAS

So if the average revenue (or value) per click of a keyword is $8 and your target ROAS is 500%, then the value per click of that keyword is $8 / 5 = $1.60. Again, you could set you max CPC a bit higher than this amount.

To speed up your bidding based on one of these formula’s above you can make these changes with bulk uploads. Using this feature, you can:

  • Download a spreadsheet with the right keyword data.
  • Add a column in Excel that calculates the right bid by using the one of the formula’s above.
  • Copy and paste the calculated value per click as values in the Max CPC column.
  • Manually review the new max CPC’s, especially the extremes (highest and lowest max CPC’s): feel free to adjust these based on your own insights and CPC limits.
  • Upload and apply the changes in AdWords.

I wrote a post 5 years ago specifically about this time saving feature in combination with AdWords Editor, you can read it over here for more detailed instructions: Align your bids with your CPA within 5 minutes using AdWords Editor and Excel.

Calculating bid adjustments

How to calculate your bid adjustments for mobile, locations and within ad scheduling has already been covered in the Campaign Settings & Bid Adjustments part of this series.

Nevertheless, it’s important enough to shortly recap and update this topic in this post.

First of all, the most important bid adjustment (that most people get wrong): the mobile bid adjustment. Google provides a clear formula and example in their Bid Like a Pro guide:

adwords-mobile-bid-adjustment-formula

Why most people get this wrong, is because they don’t include the full value of mobile which can include more cross-device conversions, phone calls, in-store visits and app downloads. Google’s example clarifies this point:

adwords-mobile-bid-adjustment-example

Location and time bid adjustments are easier to calculate, by comparing the specific location or time efficiency with the campaign average.
I’d not recommend to compare it with the campaign goal (as Google does), as you’ll bid towards your goals at the keyword or ad group level, and bid adjustments are just a way of compensating for the deviation from the average.

For CPA targets:

  • Location bid adjustment = (avg. campaign CPA / specific location CPA) – 1
  • Time bid adjustment = (avg. campaign CPA / specific time or day CPA) – 1

For ROAS targets:

  • Location bid adjustment = (specific location ROAS / avg. campaign ROAS) – 1
  • Time bid adjustment = (specific time or day ROAS / avg. campaign ROAS) – 1

For more details about these calculations, check out How To Determine Your Mobile & Geo Bid Multipliers For Enhanced Campaigns by Benjamin Vigneron and the mentioned Bid Like a Pro guide from Google.

And remember: once you use the target CPA or target ROAS flexible bid strategy, you don’t need to apply these bid adjustments anymore in the selected campaigns, as these bid strategies will factor in real-time data on device, location and time to adjust bids. Any bid adjustments you may have will be overruled by these target CPA or ROAS flexible bid strategies, with the exception of a mobile bid adjustment of -100%.

Using filters and rules

Not all your keywords have conversions and even those that have may need a closer look after applying the formula’s above.

That’s where filters and rules come in. I mention these 2 together as rules are essentially filters with a specific evaluation period and action set attached to them.

There are many ways to filter your keywords and attach bid rules to them, but my favorite framework when it comes to keyword filters and bid rules has to be Chris Haleua’s 3×3 Performance Segments. In this framework, you start by dividing your keywords in 3 main segments:

  • Keywords with conversions
  • Keywords with clicks, but no conversions
  • Keywords with impressions, but no clicks

You may also have (many) keywords without impressions because nobody is searching for them, and there is 1 simple rule for those: delete them. There’s even a script to help you do that. Or as Larry Kim said: Attention Keyword Hoarders: You Need to Delete 98% of Your AdWords Keywords.

Back to keywords with data: you complete the framework by applying the following sub-segments to the main segments above:

  • Volume leaders: keywords with many conversions, clicks or impressions
  • Efficiency threats: inefficient (too high CPA/CPC or too low ROAS) or low quality keywords
  • Visibility Struggles: keywords in low positions

Once you combine these segments, you get a beautiful framework that can help you quickly zoom in and take action on the different kind of keywords. With filters and rules, we’re mostly interested in the outliers we need to take action on.

To get you started with the 3×3 performance framework (including the cool names for each segment) I’ve created the visuals below:

chris-haleua-3x3-performance-segments

And to dive deeper into this framework and related materials, I’d highly recommend the following downloads from Chris Haleua:

  • Note: These downloads have been taken offline.

With the last Excel download you can even create a fancy 3×3 performance segment bubble chart that will surely impress your boss or client:

chris-haleua-3x3-performance-segments-bubble-chart

Once you get familiar with a set of filters that work well for your account, you can set up automated rules whenever you find yourself taking the same action under the same circumstances. The Bid Rule Playbook and Template Library downloads above should give you more than enough inspiration to create your own rules.

Bidding with AdWords Scripts

Not every ‘if then’ rule you can think of, can be translated into an automated rule. Besides, automated rules can’t run more often than daily and can’t use data outside of AdWords.
This is where AdWords scripts come in. To give you an idea of the possibilities of scripts when it comes to bidding, I’ve assembled bidding related scripts from the resources below, where you’ll also find the instructions to implement these scripts.
You don’t need to be a coder to implement existing scripts, just read the instructions carefully and copy-paste in your own account.

For a good introduction to scripts, I’d recommend these 3 presentations at SlideShare:

And If you want to create your own scripts, you’ll obviously need to be able to write JavaScript, which you can learn at Codecademy, W3Schools, Udemy and Russell Savage’s Beginner’s Guide to JavaScript You Should Know For AdWords Scripts.

Google Developer Site:

Russell Savage, FreeAdWordsScripts.com:

Daniel Gilbert, Brainlabs:

Frederick Vallaeys, Optmyzr:

If you’re interested in many more scripts to automate more than just bids, you’ll find plenty of free scripts at these resources:


Bidding using the Bid Simulator:

For your highest volume keywords, you may want to check out the bid simulator data every once in a while, as it’s the only place where you’ll get an estimate of what you can expect at different bid levels.
This is based on the last 7 days of data (including Quality Scores, competition and search volume), but doesn’t necessarily predict the coming 7 days.
If you have no reason to believe the coming 7 days will be very different in terms of competition and search volume, it should give you a good estimate of what to expect for each bid.

Let’s use the screenshot below as an example:

adwords-bid-simulator

For this keyword, we can calculate the following based on the bid simulator data above. If we would increase our bid from €0.25 to €0.50, this would lead to:

  • An increase in average CPC from €0.15 to €0.30 (100%)
  • An increase in clicks of 62%
  • An increase in CTR from 1.59% to 1.71% (7.5%)
  • An increase in cost of 214% to €3,490
  • An increase in revenue of 60% to €10,100
  • A decrease in ROAS from 568% to 289% (€10,100 / €3,490)

This can help us in selecting the optimal bid: would we rather keep the current revenue of €6,300 at a ROAS of 568% or would we rather increase our revenue to €10,100 but half our ROAS to 289%. Or any of the scenario’s in between?
Please note that you may not see the conversion and revenue estimates if there isn’t enough conversion data in your account. In that case you could multiply the click estimates with the expected conversion rate (and average order value) for that keyword yourself.

Also note that real profit maximization happens when you take incremental cost per click (ICC) into account instead of average CPC. For an introduction on this topic, watch the classic Google AdWords Bidding Tutorial by Hal Varian. If the second half of this video seems a bit puzzling, I’d highly recommend reading Averages Lie! Bid Simulator and Incremental Marketing by George Michie for a deeper understanding, thanks to banana stands (as there’s always money in the banana stand).

Combine your newfound ICC knowledge with the estimates from the bid simulator and Tenscores’ free Bid Optimizer tool and you can find the ultimate profit-maximizing bid!

Obviously, this whole process takes quite some time, so it’s mostly worthwhile for your highest volume keywords.

  1. Automated bidding with AdWords flexible bid strategies

I know, some people are fundamentally opposed to bid automation. And they especially have a hard time trusting Google to manage bids in the advertiser’s best interest.

But if they have thousands of keywords generating hundreds or more conversions a month, and they’re still bidding by hand, they’re missing opportunities.

Used correctly and under the right circumstances, AdWords flexible bid strategies can improve your results and save you lots of time when it comes to bidding. I’m talking about the target CPA and target ROAS strategies here as the other strategies are awareness-based.

You can use these strategies once the campaigns or ad groups you want to apply it to have received 15 conversions in the last 30 days. However, automated bidding works (much) better when you have more monthly conversions.

How many monthly conversions you ask? Well, Google has recently published conversion thresholds in their Setting Smarter Search Bids guide. This guide is a must read to fully understand how automated bidding works within AdWords.

And for the thresholds, I’ve summarized them in the table below:

bid-strategies-conversion-thresholds

As you can see, you need more monthly conversions for the target ROAS strategy and Google admits that target fluctuations and learning periods aren’t optimal below the 60 or 100 conversions a month.

If you’re below these thresholds, you could still benefit from Enhanced CPC (ECPC) and combine this with one or more of the manual solutions mentioned above. Or try your luck with target CPA or target ROAS, knowing it needs more learning time and target fluctuation is higher.

Once you’re above 100 non-branded conversions a month, it can work wonders and save you lots of time calculating and setting bids and bid adjustments.

There are 3 reasons why it can beat you at bidding, as you can read in the Setting Smarter Search bids guide:

  • True auction-time bidding: bids are set and signals are evaluated for each individual auction.
  • Adaptive learning at the query level: when keyword-level conversion data is scarce, it uses and learns from query-level data across your account.
  • Richer user signals and cross-signal analysis: when determining bids, many contextual signals (like user location, browser, device, etc.) are evaluated, including how certain signals work together and influence conversion rate.

The picture below shows the signals AdWords evaluates when determining bids at auction-time:

contextual-signals-bidding

So if you have enough recent conversions in your campaign(s), it’s definitely worth trying out the target CPA or ROAS flexible bid strategy.

However, always keep the following in mind when automating your bids:

  • Keep monitoring your results and bids, especially when seasonality plays an important role. If you expect sudden changes on your website, your stock or external events that influence search or conversion behavior, it’s probably best to temporarily revert to manual bidding.
  • Realize the algorithm needs to re-learn after every change you make in the selected campaigns or ad groups. So minimize changes if possible, but obviously make changes whenever necessary, like testing new ad copy and adding negative keywords.
  • Read these 5 Reasons Why You Shouldn’t Entirely Automate Your PPC Bid Management by Andrew Lolk for more things to look out for when automating bids.

For more tips on how to best use flexible bid strategies, read 7 Advanced Tips To Automated Bidding Success On AdWords Search Ads by Matt Lawson and How To Resolve Issues With Flexible Bid Strategies In AdWords by Christi Olson.

  1. Automated bidding with 3rd party technology

If your account has more than 500 monthly non-branded conversions, and especially if you’re advertising on multiple search engines or channels as social media and display, then it’s definitely worth investigating a third-party bid management solution.
Such a platform could save you many hours and improve the performance of your campaigns at the same time by providing 1 interface to manage your bids, reports and attribution across channels and campaigns. Obviously, each platform has its own algorithms, rules and capabilities and it’s important to make sure you select the one that best matches your needs.

Below you’ll find a list of well-known platforms to manage your paid media. Most of these platforms are targeted to enterprise advertisers spending $50,000 – $100,000 a month or more on paid media.
However, a few also have solutions for mid-market advertisers (let’s say between $10,000 and $50,000 monthly ad spend), those are indicated with *).

Most of these of these platforms charge on a percentage of ad spend model, usually somewhere between 1% and 5%, with the percentage decreasing as the spend increases. Next to this, most of them also charge an initial setup fee.

If you’re considering to use one of these platforms, I’d highly recommend reading Enterprise Paid Media Campaign Management Platforms 2015: A Marketer’s Guide from Digital Marketing Depot. In here, you’ll find 50 pages of analysis and advice about these platforms that can help you select the one that’s best for your needs. Another highly useful paid resource is Econsultancy’s buyer’s guide for PPC Bid Management Technology 2015.

It’s important to realize that all these platforms evolve quickly, so their capabilities could have changed by now, compared to when these reports were released. So it’s always worthwhile to ask the platforms about their most current and upcoming capabilities.

There you have it! Probably the longest post ever written about AdWords bid management. I’m curious to hear your thoughts in the comments and hope you’ve found it well worth your time!

 

Bid Management: Your Audit Checklist

All Advertisers:
checkboxHave you ensured that the relationship between search queries, keywords, ads and landing pages have been thoughtfully managed and optimized, before investing heavily in bid management?
checkboxDo you bid towards a profit-driven efficiency target, based on the value per click?
checkboxDo you track all the value that your campaigns deliver?

For lower volume accounts and campaigns: 
checkboxDo you make optimal use of bulk uploads, bid adjustments, filters, rules, scripts and the bid simulator?

For accounts or campaigns with more than 100 monthly non-branded conversions: 
checkboxHave you tried the target CPA or target ROAS flexible bid strategies?

For advertisers with more than 500 monthly non-branded conversions that also advertise
significantly on other search engines, social media or display networks:
checkboxDo you work with one of the established third-party platforms?

This is a guest post by Wijnand Meijer, Sr. PPC Strategist at iProspect Netherlands, an online media agency based in Amsterdam. He created his first AdWords campaigns in 2006 and is currently helping advertisers and coworkers alike to get their Paid Search to the next level.

Opinions expressed in the article are those of the guest author and not necessarily Certified Knowledge. If you would like to write for Certified Knowledge, please let us know

Don’t Fall for this Bad Advice! The Worst Quality Score Advice We’ve Heard

Last year I was at a conference and a speaker go up and said; I’m slightly paraphrasing:

Did you know that an ad group name can be 255 characters long? (I didn’t actually know this ). Google makes it that long so you can write descriptive ad group names to tell them what your ad group is about – you should always use all the characters as this will help your quality score. Also, campaign names can be long as well, you should use all the characters possible for your campaign names to increase your quality scores.

I speak at a lot of conferences and I attend many sessions each year. I don’t always agree with other speakers, but everyone is entitled to their own opinions – so unless it’s a panel Q&A; I won’t contradict anyone – I just might give a dissenting opinion.

Quality Score Myths

Until this time. As many audience members were glancing my way and looking quizzically as is this true or not; during the Q&A – I set everyone straight that this does not matter. Naming conventions are for you – not Google. I had hoped never to come across something like this in the wild – until this week.

I was doing an AdWords account audit and upon logging in – I was just confused by their organization. Every campaign included just one ad group. Every campaign and ad group name was more than 200 characters long of random words that reminded me of keyword stuffing organic terms back in 1998. I asked the advertiser what was going on with the naming conventions and they told me that they were told it would help quality score.

To save everyone a lot of headaches,  I thought we’d share some of the worst quality score myths we hear.

Thank you to everyone on #PPCChat who also contributed some ideas.

Put your keywords in your metatags to increase your quality score

I heard this version all the time. Put your keywords on your landing page, title tags, meta tags, alt image tags, within your video, in your video text, write them in flash, etc. I’ll also hear that having the keyword on the page is so good; but can lower conversion rates if the page is too long that they use white text on white background to not distract the visitor.

Having a keyword (or something related) on the page as it fits naturally can be useful as that’s what the user expects to see. In fact, there are other words a user expects to see on the page as well that should be there. Are you selling a good? You probably have a price, selection, and shipping information. Are you in mortgage lead gen? You probably have the word rate, calculator, refinance, and mortgage on a given page.

Those words weren’t written on the page for Google – but for the user. Write your pages for the user. If users are getting to your site, not bouncing back to search results, you’ll probably get good landing page quality scores, assuming you’re not breaking other policies, the page loads within a reasonable time frame, and that the page works on the devices where the ads are showing.

Use Popups on Landing Pages to Reduce Bounce Rates

I see this a lot in the VSL (video sales lander) space. There’s a popup that’s hard to get rid of so that if a user wants to leave your page right away, they can’t as they have to close the popup first. This popup is generally restricted from showing to bots and Google IP addresses, and if often not even added until the page has been approved, to increase quality score.

The only thing this type of page is going to do is eventually get your page disabled or for repeat offenders – banned from PPC.

Put AdSense on Your Site

This was a common suggestion about 10 years ago, but I still hear it on occasion. AdSense is the publisher platform for the Google Display Network.

The reasoning goes something like this: Google makes money when AdSense ads are clicked. By adding AdSense ads to your site, Google will make more money so they want people to go to your page over other advertiser’s and will thus reward you with a good Quality Score.

This is terrible advice. If anything, Google has punished advertiser’s for showing too many ads above the fold. While arbitrage still exists on the web; however, its usage has declined significantly since it’s heydays of 2002-2007.

Bid to the Top Position

The ads in position 1 have the highest CTR; make Google the most money – so by being in position 1, Google will reward you with a great quality score.

Now, there is a shred of truth to this myth – but not in the way people think.

Google normalizes CTR by position to determine your expected CTR; so higher positions don’t naturally get a higher quality score than ads further down on the page. However, we will see ads that as they increase in position get higher CTRs than were expected by that position, and thus QS can increase. In addition, Google isn’t always great at normalizing top vs side, especially in industries where the top CTRs often break 15%. Lastly, there are ad formatting items that come into play only in the top positions, such as extended headlines, sitelinks showing, etc.

So in some cases, you might actually see a higher quality score in the top positions than you do on the side of the page for the same ad. However, other times there’s no difference at all (or your quality score can also drop as you didn’t hit the expected CTRs) and you can quickly lose money.

So it’s not that being in the top positions gives you a higher quality score – its that if in the top positions your ad gets a different type of CTR than on the side, then raising your position can increase your quality score.

Prove Your Willing to Spend Money: Raise your bid to $1000 or your Budget to $1,000,000

I hear both the advice of maxing out your bid or your daily budget independent of each other – but the overall sentiment is the same – if you prove to Google you’re willing to spend more – then you’ll get a good quality score.

This is totally false, and it would actually destroy the auction integrity. The auction is setup so that a small account can compete with a large one. The small account can’t spend as much each day and won’t get as much traffic as the large one, but when both ads show at the same time – the small account can compete with a large one. There is no benefit to setting very high bids or daily budgets.

In the ‘dumbest mistakes ever’ category, I saw an account that tried this. They were in the legal space where the CPCs often exceed $50-$200 and they didn’t realize people were really willing to bid $250-$350 on occasion. They raised their bids to $1000, didn’t have great quality scores, and had their CPCs go from $15 to $300 in a single day. That was a costly mistake. If you raise your bid and/or budget – you can spend it – are you willing to spend $1,000,000 a day?

A High PageRank will increase your Quality Score

In 2005, I received one of the strangest phone calls that an agency can receive (I did get an email that they wanted to chat – so this wasn’t out-of-the blue, but the email didn’t prepare me for the conversation), it went something like this:

Hi, we’re an agency that has about 30 clients doing SEO and PPC for us. Google just told us that we’re not allowed to run any PPC for our clients any longer. If we stop buying PPC for them, then we’ll lose a lot of SEO business as well. We talked to Google and they suggested to us that if we run our PPC through you, and you manage the accounts, that they’ll let our clients continue buying PPC – can you help?

Google not only banned the agency – they banned the agency’s clients from buying PPC. I was surprised that Google told them they had to use us (we did have an amazing relationship with Google at the time as one of their largest resellers and clients). I had to find out why all their clients were banned.

The agency was:

  • Buying links to landing pages
  • Getting landing pages approved that passed Google’s guidelines and then using sneaky redirects to send the users to other pages
  • The agency and their clients had been repeatedly warned that if they kept up these practices they would be banned
  • The agency thought Google was bluffing
  • Google doesn’t bluff about AdWords

When pressed about why they would do this; the agency told me it was to get good Quality Scores. They needed higher pageranks to increase the quality score because their clients were all lawyers or plastic surgeons and the CPCs were too high.

An entire agency banned over a quality score myth: A high PageRank does not lead to a great quality score. In fact, you can have your site banned in the organic results and still get 10 quality scores.

Link to Google from Your Landing Page

For this example – I choose the link to Google. You could replace that with any spammy 2000 SEO tactic.

Your spammy SEO tactics do not affect quality score. Google figured this out already. Remember penguin, panda, hummingbird, and all the other Google updates? Google didn’t forget update Adsbot-Google with its organic crawl technology.

Create One Keyword Per Ad Group & Use the Keyword Twice in the Ad

It’s highly debatable if you should use all SKAGs (single keyword ad groups) in your account to begin with, but making them just for quality score purposes is definitely not one of the ‘pros’ for that type of organization.

Including the keyword in the ad is useful if it helps CTR (and thus quality score) or increases relevancy; but you can achieve quality score 10s without having a keyword in the ad.

The overall rule for ads when it comes to quality score is simple: test different variations and see what does best.

There’s not a ‘rule’ for if you do this exact one thing every time you’ll get a high quality score.

This leads us to our next myth…

Always Use DKI – DKI increases Quality Score

This myth has a shred of truth in it – but not in the way many people think.

First off, in your account, you see your quality score. This is not what’s used in the auction – it’s just a mashup of a lot of data to give you an overall indication of your quality score and keyword’s health. It’s argued that in the real time auction, having the keyword in the ad can slightly increase the real time quality score – and since you’re using DKI – the keyword is always in the ad – and thus a higher quality score.

Let’s do some quick debunking:

  • The keyword put into the ad is not the search query – it’s the keyword from your ad group that caused the ad to be displayed. These might not be related words – especially in the case of broad match usage.
  • If your keywords are over 25 characters – your keyword won’t appear in the headline (although, sometimes Google does put up to 28-29 characters in the headline).

In reality – this is how you should think about DKI’s relationship to quality score:

  • If using DKI increases your CTR; then your QS can go up.
  • This increase in QS has nothing to do with the fact your used DKI
  • This increase is because your CTR decreased

If you use DKI and your CTR goes down, as your ad is less compelling that your previous ad, then the use of DKI can lower your quality score. The act of using DKI is immaterial to quality score. The resulting metrics from your (with or without DKI) is what affects quality score.

Using Every Match Type will Increase Quality Score

The quality score you see in your account is based upon when the search query was the same as your keyword regardless of the match type. If you have match types in different ad groups with different quality scores – it is due to the ad differences or some other random factors. Adding all the match types doesn’t increase quality score – it just gives you more transparency into how your keywords are doing with different sets of ads.

Delete Low Quality Score Keywords to Improve Your Overall QS

Theoretically, there is no account quality score. A Google rep doesn’t have this number. In practical terms, there is an account level score. You can calculate this yourself using weighted quality score data.

One keyword’s effect on another keyword’s quality score is so low – it might as well be zero.

There’s a bigger issue at stake with this myth: If a keyword is profitable – who cares if Google gives it a low quality score?

I can’t count the times I see someone just delete all low quality score keywords to make Google happy regardless if the word is profitable or not profitable. If Google gives me all 10 quality scores but I can’t make a sale – I’m going out of business. If I have all quality score 3 keywords and I’m making money – I might try to improve my quality scores; but that’s because I’m making money and am still in business to try and take that step.

Profit is more important that quality score!

Old Accounts Get Better Quality Scores

Once upon a time, you’d see AdWords accounts for sale because they were old. Not because they were good; but because they were opened in 2002 or 2003.

The older the account and the higher the gross spend – the higher the sale price.

Luckily, this seems to have passed us buy. Don’t buy someone’s old PPC account and expect it to generate high quality scores just because it is old.

Making Changes to Your Account Will Lower Quality Scores

Have you ever been afraid to move a keyword? Did reorganizing your account make you shudder, not because of the time involved, but because you’d lose your history and your quality scores would suffer?

If yes, you managed accounts in the mid-2000s. This was a real issue once upon a time.

That time has passed.

Use.A.Period Not a Space To Separate Words

There was a time when Google was not forthcoming what separated a word. You’d see ad groups like:

  • I’m a keyword
  • Im a keyword
  • imakeyword
  • im.a.keyword
  • I’m.a keyword
  • im-a-keyword
  • I’m_a_keyword

Google has since become much more transparent

Using symbols with your keywords

You can use two symbols, ampersands (&) and accent marks (á), in your keywords. We’ll consider keywords with these symbols as two different keywords, like sidewalk cafe and sidewalk café.

Here are some of the symbols that our system doesn’t recognize:

  • Ignored symbols: You can add periods (.) or dashes (-) to your keywords, but we’ll ignore this punctuation. That means we’ll consider the keywords Fifth Ave. and Fifth Ave, for example, identical keywords.
  • Invalid symbols: You’ll see an error message if you add keywords that contain certain symbols. Some of the symbols that can’t be used in your keywords are: , ! @ % ^ * () = {} ; ~ ` <> ? \ |
  • Search operators: We’ll remove the “site:” operator from your keywords. That means if you add the keyword [site:www.example.com dark chocolate], we’ll consider it the same as [dark chocolate]. If you’re someone who uses advanced search operators on Google, you might want to know how those operators work with your keywords. If you add a particular search operator to your keywords, AdWords will ignore it. For example, if you add the search operator “OR” to the keyword dark chocolate, like OR dark chocolate, AdWords will ignore the “OR” command and target your ads with the keyword dark chocolate.

Source – AdWords help.

Your TExT CASinG ChanGES your QualitY ScORe

Google didn’t use to give you poor quality scores and let you continue to advertise – they would disable your keywords and you could never use that keyword again in the account. This was a real problem in 2004-2006 when your best keywords would fall below a 0.5% CTR and you knew they were about to disappear. There was a trick to the system – you could write the word in a different casing and Google would re-enable that keyword.

Google’s grown up since then – the old tricks don’t work anymore.

Google Hates Affiliates – Obfuscate Your Affiliate URLs to get Higher Quality Scores

Google hates bad advertisers, sites that shouldn’t exist, and sites that put a barrier between a searcher trying to find an answer and the answer itself.

Google doesn’t care if it is an affiliate, a publisher, or an advertiser in the way – if you’re slowing down a searcher – your bad news.

Did you realize that Expedia, Orbitz, TripAdvisor, Hotels.com, etc are affiliates? They are big affiliates, but they are still affiliates. The largest advertisers on Google are often affiliates.

Google doesn’t care if you’re an affiliate. If you’re providing value to the search ecosystem – then you’re OK.

What Other Bad Advice Have You Heard?

Have you heard some other bad advice recently that you want to warn people about?

Just let us know in the comments.

Free Webinar: Improve Your Landing Page Conversion Rates Through Ad Testing

Did you realize that your PPC ads don’t bring in conversions?

Your PPC ads ensure that you are brining the correct type of traffic to your landing page that your website can then convert.

As your PPC ads help to qualify the traffic and are highly measurable, you can use ad testing to help design your landing pages so that the user’s have a seamless experience between your ads and the landing page. This type of testing can also help you create a consistent offer between your ads and landing pages.

On Tuesday, November 10th at 1pm EST / 10am PST myself (Brad Geddes) and Formstack are going to do a joint webinar that examines how to improve your landing page conversion rates via Ad Testing.

You can register for the webinar here.

We hope to see you there 🙂

Google’s Boom is Over

While Google’s homepage hasn’t, bar for a few tidbits, changed in the last years, the behavior of online searchers has dramatically changed. When Google originally started their services in 2001 there wasn’t many websites and products to search for – at least compared to 2015.

For this reason users were happy to find a matching website or product that solves their problem.

But users have changed their ways and their expectations have changed. Users are now looking for the best match. Any they know that in many cases it isn’t always on Google.  Online searchers are developing new habits counter those we have gotten used to in the past.

What does this mean for Google and us search marketers and advertisers?

 

Forgetting the Big Picture

Day for day one toils away doing whatever one must to increase the performance of one’s advertising, website or even internal matters like sales and production.

In some cases this is done in adherence with strategic plans laid out after careful research of the past. In other case actions are capriciously undertaken, based on the last webinar or the latest post read on inbound.org that contained some good ideas.

Either way, one can forget to kept an eye on the data of the macro environment in which ones business model resides.

Macro environment is defined as: The conditions that exist in the economy as a whole, rather than in a particular sector or region.

In this case the macro environment in question is Search Engine Marketing on Google.

(In Germany Google has an apparent market share of over 95%, at least when looking at the main engines Yahoo, Bing and a few local ones, so for us it is apparently the only relevant search engine – more to that in a bit)

And if I would reduce all the lessons of life in last 20 years into one, then it would be that:

Changes is inevitable and in many cases they are just around the corner.

Internet, Social Media, tools, technology, demographics, politics etc….. Things are in a constant rush for change, nothing stays the same for a long time.

The same can be said of markets – markets always change – nowadays more than ever.

So it makes sense to take a look at the macro data once in a while. Google Trends is a great tool for this.

(I wouldn’t swear on the accuracy of Google Trends, but for the sake of thought provocation and fun, let’s assume that it is more right than wrong)

Additionally, what the internet has come to mean so much for so many businesses, it is well worth the time.

 

Online Advertising Is Relevant For Most Businesses

Ok, for starters, let me regurgitate a few headlines of the latest business magazines and internet blogs and hopefully prove a point:

The past decade online marketing has grown and become for most business owners a serious topich and a high-priority on their to-do-list.

Proof: spend in online marketing has been climbing and climbing. Google, the biggest advertising company in the world, increased its yearly turnover sky high to 66 billion since 2001. Take a look at the graph.

 

Google Turnover Keeps Climbing Since 2002

1-Google-Turnover-Since-2002

Source: www.statista.com/statistics/266206/googles-annual-global-revenue/

That is of course incredible. Google is by now one of the biggest companies worldwide. In online advertising, they are by far the strongest company.

And to just to make another case in point: E-Commerce turnover in 2014 has increased to a record 304.91 billion in the US and 43,6 billion in Germany  – the country I call home.

The boom in Google’s turnover and that in E-Commerce businesses doesn’t seem to be ending soon.

But that doesn’t mean that online advertisers and website owners can expect unlimited, never-ending growth.

Our income is not tied to Google’s or to any online retailers for that matter.

(I don’t want to know how many online businesses are a far cry from being successful or just vaguely profitable. )

 

Search Volume Growth Rate Slows Down

Interestingly, yet understandably, the growth rate of search queries is not increasing. It is still decent growth but the volume is normalizing at a high level.

Google searches per year:

Steven Ruesch: -Google-Searchqueries-per-year

Source: internetlivestats.com/google-search-statistics/

Google Search Volume Growth Rate (in %)

Steven Ruesch: 3-Google-Search-Volume-Grow-Rate

Source: internetlivestats.com/google-search-statistics/

 

This of course represents the overall numbers worldwide. And considering the increased use of bots to collect data on Google, manipulate CTRs and so on, this chart might even look very different.

Another interesting graph is the search volume for “Google” on Google according to Google Trends in the US:

Steven Ruesch: 4-Navigational-Searchqueries-Google

Considering the growth level of Google’s turnover the numbers are somewhat surprising or even disappointing, but understandable. (Remember everything changes – even markets)

But do these overall figures apply to all markets?

No.

 

Search Volume Changes From Business to Business

 

In some markets like dentistry (US) we have a steady growth rate exceeding the average:

Steven Ruesch: 5-Google-Trends-Dentists

 

Search queries (US) for insurance on the other hand has been declining – does that mean insurances are being sought and bought less often? No it doesn’t. I added a complete explanation at the end of the post.

Steven Ruesch: 6-Google-Trends-Insurance

 

So has queries for holidays (US):

Steven Ruesch: 7-Google-Trends-Holiday

 

While people have been searching for smartphones (US) incredibly often:

Steven Ruesch: 8-Google-Trends-smartphones

 

Ipads were being searched a lot, not so much anymore:

9-Google-Trends-Ipads

 

Movie streaming is a nicely growing market:

10-Google-Trends-Movie-Streaming

As you can see, the figures change from market to market.

But one can safely state that the exponential growth we had is in average over.

We have a high level of traffic that is being shared by many websites. But in many cases, it isn’t growing.

This situation as defined in the product cycles is the 3rd phase, which is the Maturity stage. The full list of stages are: product introduction, growth, maturity and decline.

In this stage we have websites competing for the same valuable traffic, where sales growth rates are diminishing and the primary strategy is to defend ones market share. – Sound familiar?

 

The Role of Search as We Know it Will Drop in Importance

Yep that is my earnest opinion.

The days of one search engine for all types of searches are ending.

Just like any other product or advertising media for that matter (even when newspapers try to negate the laws of nature), there comes a time when other better forms of a product emerge that replace that of contemporary ones.

This always happens – it might be a better search engine or just the rise of hundreds of specialized search engines as we are already seeing with Amazon for products, or citations or review sites for local businesses.

It is only a matter of time.

To illustrate my point, let me go back to the topic of insurances and their search volume in Germany:
(For those who want to verify the figure, the German word is Versicherung)

11-Google-Trends-Versicherung

I think you will agree, the figures are declining or at least not growing.

Now compare that to the search volume for Check24 – the market leader for comparing insurance policies in Germany:

12-Google-Trends-Versicherung-vs-Check24

Of course this volume isn’t solely navigational search queries from people looking for the check24 website, but it does indicate the popularity of their service within the German market. The main search query according to Google trends is: “check24 versicherung” (German for insurance).

In this case we have users who are bypassing Google as the search engine while searching for an insurance.

Users who have gotten so proficient at search that they are mastering the ability of using more than one engine.

(No Google, users don’t want one complex search engine to fit all needs. They would rather have distinct services with their own set of features)

I am sure you can think of several cases where we have this situation.

It also raises the question (which has been raised numerous times already), whether market share comparison of search engines is valid when only including the usual suspects: Google, Yahoo, Bing and the other odd few and ignoring giants like Amazon or even services as those similar to check24, who are technically also a search engine.

As an advertiser the interesting question to pose is: When does it make sense to seriously start reallocating budgets to other forms of advertising? (This is already happening in the USA, but not so much in Germany)

This point won’t come soon for most micro environments within the Google cosmos but I have seen situations where advertisers are legitimately leaving or shunning Google due to ridiculous market conditions, looking for greener pastures elsewhere.

This point will arrive where Google loses it search market clout. The question is just when and how exactly.

Or does Google have an ace up their sleeve, whereby they introduce something so revolutionary, that it redefines the market in their favor.

Search engine advertising and search engine marketing have been so good to us, that we could forget to heed the laws of markets and fail to see the overall picture.

But we know better, don’t we?

Would love to hear your thoughts on this issue.

Catch Up With the Latest in PPC Thoughts on AMA

I recently did an ‘Ask Me Anything’ on Reddit.

We covered a lot of information from branding to tools to career development and lots of subjects in-between.

You can view the entire thread (and chime in) here: Brad Geddes AMA.

I hope you enjoy the thread – it took several hours to write everything to cover this many topics.

The Complete AdWords Audit Part 12: Google Shopping Campaigns

This is a continuation of the AdWords Audit Series. You can see previous parts here: Introduction, Goal setting, Measurement, Campaign Settings & Bid Adjustments, Ad Extensions, Impression Share & Auction Insights, Quality Score, Account Structure, Keywords & Match Types, Ad Copy, Testing and the Lin-Rodnitzky Ratio.

 

If you’re retailer, Google Shopping is one of the most important things to get right in your paid search program. And if you’re not, it’s usually impossible to promote your products or services with Product Listing Ads (PLAs).
So for those of you who aren’t managing (or planning to manage) any Shopping campaigns: I’m afraid you might as well skip this part of the audit series. However, the next part about bid management should appeal to all advertisers and agencies, as should all previous parts you can find above 😉

And for those of you that want to know all about how to achieve success with Shopping campaigns, this post is divided in the 4 following parts:

  • Why use Google Shopping? Because it works, it’s growing and it keeps getting better.
  • How to get the most out of Shopping Campaigns? It all starts with a great product feed. Next to that, I’ll talk about campaign structure and optimization best practices.
  • Shopping resources and tools. Especially when it comes to Shopping, there’s nothing wrong with some external help of specialized agencies or a data feed management tool to get the best results from PLAs.
  • Your Audit Checklist. As with all parts of this series, the post ends with a checklist to use for your own account(s).

As Simon Sinek says: let’s start with why. Why is it that important to get the most out of Google Shopping?

It works
Although you have much less control over which product will show for a specific search query, almost every Shopping campaign I’ve seen outperforms its non-branded text-based counterparts.
In the first place, it’s probably because a picture is worth a thousand words. And that’s a lot more than the 95 characters you get in a text ad.
Secondly, not only does the user see the actual products before clicking on them, prices are also shown next to each product.
So whatever product the user decides to click on, (s)he has been able to easily compare the products and prices before the click. This could make it a much more qualified click than a click on a less specific (or harder to compare) text ad.

Just take a look at the SERP below after a search for ‘juicers’:

Not only are the PLAs in the top right (they also appear in top positions above the organic results by the way) the most noticeable part of the SERP, the same advertiser can easily show multiple PLAs, as Walmart does in the example above.
If the same advertisers also has a text ad for that keyword, he has taken up even more of the coveted real estate on the SERP.

Shopping campaigns outperforming their non-branded text-based counterparts is not just my personal experience, it also clearly shows in Merkle | RKG’s Q4 2014 Digital Marketing Report, where we see the following PLA Performance vs Text Ads:

 

The most interesting comparison is for non-branded performance, as PLAs mostly show for non-branded queries (unless you’re a manufacturer). Looking at the yellow charts above we see the following differences:

  • PLAs had a 5% higher non-branded CPC
  • PLAs had a 130% higher non-branded CTR
  • PLAs had a 9% higher non-branded ROI
  • PLAs had a 30% higher non-branded Conversion Rate
  • PLAs had a 12% lower non-branded Average Order Value

Obviously, your PLA performance may vary, but it’s rather exceptional to see PLAs perform worse than non-branded text ads for the same advertiser.

If that’s your case, and you followed all the advice below, the most common reason for underperforming PLAs is when an advertiser isn’t competitively priced while the consumers are very price-sensitive.
In those situations it’s hard to make Product Listing Ads work, as a user can directly compare prices on the SERP.

It’s growing, fast.
More and more paid search clicks for retailers come from Google Shopping. Looking at the Q4 2014 Global Digital Advertising Report by Adobe Digital Index and the mentioned Merkle | RKG report, we see the following stats:

 

Both show almost identical growth rates when it comes to ad spend on PLAs: 45%-47% YoY growth in Q4 2014.

So what does that mean for the percentage of clicks coming from PLAs for retailers?

According to Adobe, Google Shopping produced 20% of all retailer paid search clicks and for the clients of Merkle | RKG it was even higher: 30% (Google paid search clicks only).

Let me bombard you with stats once more, but possibly the most interesting one. Below you’ll find the PLA share of non-brand Google paid search clicks, per industry, also from the Merkle | RKG report:

As you can see, PLAs produce between 31% to 71% of non-branded paid search clicks for retailers, depending on the industry.

These are impressive numbers and we can only expect the importance of Shopping campaigns to grow.
If you see lower click shares coming from PLAs in your account, there may be interesting opportunities to get more of these profitable clicks by following the advice in the rest of this post.
For advertisers outside the US it’s important to note these are mostly US based numbers, so it could be lower for countries where Google Shopping is less evolved.

It keeps getting better
Since their introduction as a beta in the US in 2009, PLAs have come a long way. Just take a look at the most important updates in the table below (click for a larger, more legible version) and you’ll quickly come to the conclusion that Google is committed to make Shopping campaigns an even bigger success. In 2014 they almost averaged 1 big Shopping improvement or expansion a month.

 

So now that it’s clear that it works and that it’s growing and improving all the time, it’s time to go into best practices to help you achieve the best possible results from your Product Listing Ads.

How to get started with Google Shopping?

Just in case you don’t have PLAs running yet and you’d like to get started, this is what you need:

 

How to get the best results from Shopping Campaigns?

It all starts with a great product feed

Your product feed is the most important factor when it comes to success on Google Shopping. As Google likes to say: your feed should be as fresh, accurate and comprehensive as possible. This is what they call high-quality product data.

As you’ll see below, there are certain attributes of the feed you’ll like to optimize regularly, based on search behavior, just as you would optimize your website for SEO purposes.

So it’s best if the person or team running your Google Shopping campaigns can also make changes to the feed, or can at least suggest changes that will be implemented by someone who manages the feed.

Below, I’ll go deeper into detail about best practices and requirements for the most important feed attributes.
But first, be sure to have these 2 links in your bookmarks to refer to whenever you’re improving the feed, so you’re always using the most recent feed specification:

 

3 Required attributes to keep optimizing:

Below you’ll find the “big three” that highly influence your visibility and CTR (next to price): title, description and image.
These are the ones you’ll want to keep optimizing for better results. Hopefully, the recommended size, do’s and don’ts below will help you do just that.
This also means you should regularly analyze the search terms report within your Shopping campaigns to guide your product feed improvements and to find negative keywords.

1) Title:

Recommended size: your title can be up to 150 characters, but it will get truncated after 25-30 characters on the SERP and after 70 characters on google.com/shopping. So front-load your titles with the most important words.

Do’s: keyword-rich; clearly describe the product; include the brand (even if it’s already in the brand column); include the category (even if it also has separate columns); make it match with the product title on the landing page; use simple color names, which means converting too specific color names to one of these 11 colors, possibly adding ‘light’ or ‘dark’ before them (if applicable): black, blue, brown, gray, green, orange, pink, purple, red, white and yellow.

Don’ts: keyword stuffing; your brand name (unless you’re a manufacturer); all capitals; exclamation points; promotional text (e.g. “Free Shipping”); use of words with low search volume (e.g. SKU numbers, too specific color names).

2) Description:

Recommended size: your description can be up to 5,000 characters, but Google recommends to keep it between 500 and 1,000 characters (and 500 should be more than enough). The description only shows on google.com/shopping and gets truncated after 170 characters. So make sure the most important words are in the first 170 characters of your description.

Do’s & Don’ts: same as for title, but you have the room to use full sentences in the description. The best way would be to treat it as a very long search ad. This means using relevant keywords, mentioning features, benefits, USP’s, call-to-actions, etc. Furthermore, the following information should not be in your description according to Google:

  • Links to your store or other websites
  • Billing, payment, or sales information
  • References to internal categorization systems, e.g. “Toys & Games > Toys > Baby Dolls”
  • Comparisons to other products beyond the item you’re selling, e.g. “even better/faster/larger than X”
  • Details about other products, accessories, or similar items available

3) Image_link:

Recommended size: the link itself can be up to 2,000 characters. The image should be the largest, highest resolution, full-size image you have for the product, but no larger than 4 MB or 64 megapixels. Google recommends at least 800 x 800 pixels. Minimum size is 250 x 250 pixels for apparel and 32 x 32 pixels for other industries.

Do’s: high quality images that differentiate you from the competition; show the product clearly (white background is usually best); test different product angles.

Don’ts: low resolution images; watermarks; logos; odd backgrounds; multiple products in 1 image; stock photos everyone is using; cluttered images.

To get an idea of how you’re currently doing when it comes to the attractiveness of your titles, prices and images and of the optimization of your feed in general, you can compare your CTR to the benchmark CTR and monitor your Impression Share in the Product groups tab of your Shopping campaigns.
But first, make sure your budget and bids are high enough to ensure a decent Impression Share.


Required and recommended attributes to increase data quality

The attributes below don’t need to be optimized continuously, but need to be as complete and accurate as possible.

  • ID: your internal and unique ID for each product, max of 50 characters.
  • Link: the landing page of the product, its content should obviously match with other attributes like title, description, price, availability and image_link.
  • Condition: new, used or refurbished.
  • Availability: in stock, out of stock or preorder.
  • Price: the actual price obviously. Being competitively priced is even more important within Product Listing Ads, as users can compare prices directly on the search results page.
  • Brand: simply enter the brand name of the product. For some brands it’s best to use the most used version of the brand (e.g. ‘Tommy Hilfiger’ instead of ‘Hilfiger Denim’)
  • GTIN or MPN: the corresponding unique product identifier of the product like a UPC, EAN, JAN, ISBN or MPN number. Submit the ‘identifier_exists’ attribute with a value of FALSE if your product doesn’t have a GTIN or MPN (e.g. for custom-made products).
  • Google_product_category: Select the most relevant category from Google’s product taxonomy.
  • For apparel products: gender, age group, color, size, material, pattern, etc. are recommended attributes to add for apparel products. If you sell apparel, be sure to read the Submit apparel products guide by Google.


Attributes to customize AdWords organization & bidding:

When you create product groups in AdWords to bid on, most of the feed attributes you can use are more or less fixed: Item ID, Brand, Category (based on the Google product taxonomy) and Condition.

However, if you want to have more freedom in how you organize and create your product groups (and with that, how you differentiate your bids), you can use one (or both) of the attributes below to do so:

  • Product_type: this is your category of the item, so you can provide your own classification next to Google’s (which is fixed and according to their product taxonomy).
    Be sure to provide full strings separated by “ > “, e.g.: Home & Garden > Kitchen & Dining > Appliances > Refrigerators
  • Custom_label_0 through Custom_label_4: Use up to 5 custom labels for extra categories such as best sellers, seasonal products, promotions or for margins.

For inspiration, see the examples below from the official support page (“How to use custom labels”):

custom-labels-examples

The mentioned feed attributes above don’t include all product feed specifications, so make sure to include all attributes in your feed that are required and recommended.

Next to having a great product feed, don’t forget to:

  • Start with using test data feeds to check your data feed for errors first. Once you are happy with the feed processing results, you can upload the same file as a normal feed.
  • Upload your feed regularly in Merchant Center. At least once every 30 days, but daily would be better (and more often if needed).
  • Make sure to regularly check in Merchant Center for possible errors and missing attributes. You’ll find such updates and warnings in the Diagnostics and Feeds tab of Merchant Center.

After reading more than a 1,000 words about product feeds, it’s time to reward you with an example merchant center feed (Google spreadsheet) that should clarify and summarize all mentioned attributes and best practices above. Enjoy and feel free to share!

Campaign Structure & Optimization

Organizing and optimizing your Shopping campaigns is totally different from regular keyword-based search campaigns. As mentioned before, you really need to know exactly how product groups, bids and campaign priorities work to fully understand the different organizational strategies.

And as often in PPC, there’s not just 1 right way to organize your Shopping campaigns.

However, if you know beforehand which attribute(s) of your feed will mostly determine the bids you’ll set at the product group level, the table below can help you with your first organizational choices:

Next to the basic choices above (that don’t include the cool things you can do with campaign priorities) there are quite some other things to consider when organizing and optimizing Shopping campaigns:

  • There seems to be a 10,000 product limit per campaign in AdWords (although you won’t receive any warnings). So if you have more than 10,000 products in your Merchant Center feed, it’s probably best to organize your campaigns in such a way that each campaign targets no more than 10,000 products.
  • The order in which you choose your initial product groups is important, so draw out your structure before creating product groups. Since every product group is a subset of the former group, it’s hard to reverse your initial choices.
  • Use ad groups if you want different negatives, different mobile bid adjustments or different promotional messages for different product groups, as these are set at the ad group level.
    Realize users only see your promotional message when they hover over your product image and that they apply to all products in your ad group. So as Google says: “These messages shouldn’t be used as ad text, but rather as actionable alerts that differentiate you from the competition.”
  • Next to the Product groups tab, don’t forget to regularly use the Dimensions tab in your Shopping campaign to view performance data by all available dimensions and product attributes, even down to the individual item:
  • Be careful not to blindly raise your max. CPC to the Benchmark max. CPC. Before increasing bids for better results, make sure your product feed and structure are as good as they can be and that your daily budget is high enough. If you still miss a lot of Impression Share or if your product groups perform great and you want more volume, then increase your bids cautiously.
    Note: during peak seasons, you may want to quickly and aggressively increase bids for maximum exposure.
  • As mentioned before, regular analysis of the search terms report is essential. On the one hand to guide the optimization of the titles and descriptions in your feed, on the other hand for query mining, both negative and positive:
    • Add negative keywords for queries that are irrelevant and/or significantly underperforming.
    • If you see high-volume queries that perform well, make sure they’re also active in a regular search campaign. If not, add these as new keywords to see if you can get even more clicks from these queries by adding a text ad on the SERP.
  • Use all the ‘extra’s’ that are available in your country, such as:
    • Dynamic Remarketing: as a retailer with a product feed in Merchant Center, it’s very easy to retarget your site visitors on the Google Display Network with ads that automatically contain the products they viewed on your website.
    • RLSA: ask your Google rep if you don’t see the ‘Audiences’ tab in your Shopping campaign to enable this, as you’ll want to be able to bid differently (usually higher) for users that already visited your website.
    • Merchant Promotions: currently available in the US, UK, Germany, France, Australia and India.
    • Product Ratings: currently available in the US, UK, Germany and France.
    • Local Inventory Ads (if you have local stores): currently available in the US, UK, Germany, France, Australia and Japan.
    • Google Trusted Stores: currently available in the US, UK, Germany, France, Australia and Japan.

Advanced Shopping campaign structures

Most advertisers start to organize their Product groups based on category or brand and have just 1 Shopping campaign for all their products. And often, that’s fine.

But if you have a lot of products, sales or other factors that influence performance, you’ll want to get more advanced in your organization and bidding by using the campaign priorities setting.

In this video (just a little over 4 minutes): How to Best Structure your Google Shopping Campaigns by Merkle | RKG you get a great and quick overview of 3 campaign organization strategies and the advantages of using this Shopping campaign structure (using priorities):

A very interesting variation on the structure above that you should definitely check out is Taking Google Shopping to the Next Level by Martin Roettgerding.
In a 27 minute video (scroll down to see it), Martin shows you how to use campaign priorities to be able to bid the highest on the most specific queries and lowest on the most generic queries, even if all these queries match the same product.

The structure looks like this (taken from his corresponding SlideShare presentation):

And an example of the outcome of this campaign structure looks like this:

 

Agencies, Data Feed Management Services & Tools:

You may not have all the (optimal) resources or technology for managing product feeds and/or Shopping campaigns.
As these are essential to retailer success with PPC, you can reach out to one or more of the approved Google Shopping Partners to help you with this. Shopping Partners consist of specialized agencies as well as feed management services.

If you want more choice in tools for product feed management (besides Google Shopping Partners) you could also search for ‘product data feed management’ and check the organic search results. A safe choice would be Google’s own Channel Intelligence (acquired in 2013), as they’re obviously specialized in feeds for Google Shopping, but also many other Shopping Engines.

When it comes to tools for more efficient management within AdWords, the Optmyzr Shopping Campaign Tools could save you quite some time. Check out the demo videos on the Optmyzr website to get an idea how these work.

Just as your product feeds should be, I tried to be as fresh, accurate and comprehensive in this very long post about Shopping campaigns.
However, if you have any comments, questions, suggestions or experiences you’d like to share, please leave a comment below!

 

Google Shopping Campaigns: Your Audit Checklist

Merchant Center – Product feed:

checkboxIs your feed fresh and does it contain all products you want to advertise for, i.e. does it match with the contents on your website (prices & availability especially)?
checkboxAre all the required and recommended feed attributes as accurate and comprehensive as they can be, based on Google’s feed specifications?
checkboxAre the titles and descriptions of your feed user-friendly and do they match search behavior? Regular query mining should guide your feed improvements.
checkboxDo you use and test the highest possible quality product images that distinguish you from the competition?
checkboxDo you use the product_type and custom_label attributes to be able to bid based on your own categorization?
checkboxDo you upload your feed as often as you should, given changing prices and availability? At least once every 30 days is required, daily is usually recommended.
checkboxDo you regularly check the Diagnostics and Feeds tab in Merchant Center for warnings, errors and missing attributes?

AdWords – Structure, optimization and bidding:

checkboxDo you make sure each Shopping campaign targets no more than 10,000 products?
checkboxDoes your structure focus on product lines, profit margins and best sellers?
checkboxIn case you have seasonal items and/or flash sales: do you promote these with separate campaigns, using different campaign priorities?
checkboxDo you take full control of ad serving with all 3 campaign priorities (Low, Medium and High)?
checkboxDo you run search query reports to find negatives to exclude irrelevant or poor performing queries?
checkboxDo you run search query reports to find new keywords for your regular (text-based) search campaigns?
checkboxDo you use the Benchmark (CTR & CPC) and Impression Share insights to guide your optimizations?
checkboxDo you use the insights from the Dimensions tab to guide your optimizations?
checkboxDo you use separate ad groups whenever you want to have different negatives, mobile bid adjustments and/or promotional messages for your product groups?
checkboxDo you use all additional Shopping (related) options that are available in your country, such as Dynamic Remarketing, RLSA, Merchant Promotions, Product Ratings, Local Inventory Ads and Google Trusted Stores?

 

This is a guest post by Wijnand Meijer, Quality & Learning Manager at iProspect Netherlands, an online media agency based in Amsterdam. He created his first AdWords campaigns in 2006 and is currently helping advertisers and coworkers alike to get their Paid Search to the next level.

Opinions expressed in the article are those of the guest author and not necessarily bgTheory. If you would like to write for Certified Knowledge, please let us know

How to Work with Legal & Brand Management For Better Ad Testing

This post originally appeared on the adAlysis ad testing blog.

 
wooden gavel and books Many marketers must submit all their ads to a legal and/or corporate branding team for approval before they are allowed to use them in a PPC account. These departments can often be bottlenecks to ad testing.

If you aren’t working well with legal and/or branding; then these departments can quickly become an excuse for why you aren’t testing ads appropriately and why the account is not growing as anticipated.

In this article, we’ll explain how to work with legal and branding so that they can help you in your ad testing process and not be a hindrance.

Explain Why Ad Testing is Important

The first step is sitting down with the departments and explaining why ad testing is so important. Proper ad testing can result in:

  • Higher quality scores
  • Lower CPCs
  • Higher positions at the same cost
  • More traffic
  • Lower cost per actions
  • Higher ROAS
  • More conversions
  • More revenue

You must explain to them why ad testing is both important to your job, but also crucial for the company as a whole.

Ask for a Time Commitment

Once the department understands why you must test ads; explain to them why you need their approval or feedback. Make them seem included in the ad testing process, not just a checklist item that you are forced to endure. The more a person feels special and included in the process, as opposed to just being a barrier or paper pusher, the more involved they will be in the process.

Once the person or department understands why you need their help to make the company grow and become more profitable online, then ask for a time commitment.

Try to schedule a time every single week where the department can approve ads and give you feedback. In some cases, you might only be granted monthly hours, but if you can get a weekly time slot it will make your ad testing more consistent.

Regardless of when the review process occurs,  agree on a recurring period of time where they will approve ads and give you feedback.

At first, you might have to follow-up constantly to get them to stick to this time commitment.  However, being organized and pushing for ad approval and feedback based upon their agreed upon schedule, will help in the long run as they get into the routine of approving ads and sending you feedback.

Determine the Best Ad Format to Send them Data

Before you go to the meeting, print out ads in a few methods:

  • AdWords editor import: With this printout, you will have an Excel sheet that lists the headline, d1, d2, display, and destination URL, department approval status, and notes by column.
    • This format is easier for you to deal with since you can just upload the data once it’s approved. However, a lot of non-PPC people find this a difficult format to work with.
  • Excel/Word Doc format: In this format, you will have the ad in one column as it will appear in a search result. Then another column for approval status and notes.
    • This format is easier for a lot of non-PPC people to read as they can see all the lines at once. However, you will have to do a bit more work putting the ads in this format; and once the ads are approved, you’ll have to put it into another format to upload to your account.
  • Note: You could also put this into a shared drive, Google Docs, Microsoft Office Online, etc format so that you can just send them a link to an online document they can read and edit.
  • Note: Your company might have some specific checklists or feedback requirements for ad approval. If so, add those to the sheet ahead of time so you can show you do understand the entire process and are trying to make life easier for them.
  • It’s ok to elicit feedback on the forms. Your goal is to make this as easy as possible for the department.

Ask them which format they prefer. While this might take you a bit more time, it will ensure that the department can work easier with your data and approve more ads within the same time commitment.

Making ad approval easy for legal and branding will help ensure that they realize you are trying to work with them and the easier it is for them, and the more ads you will get reviewed by the departments.

How Many Ads They Can Approve in a Time Frame

Checklist Once you have agreed on a format, ask them how long it will take them to approve ads in that format. Bringing a few printouts of varying numbers of ads to the meeting can help them see how many ads you really do need to get approved.

Your goal here is to get a sense of how many ads you need to create in a timeframe so that you can build up a queue of approved ads for testing. Even if you only need ten new ads a week; if they can approve 100, take advantage of that time to get more done than necessary.

What you don’t want to do is have them agree to an hour a week and then send them only 10 minutes of work per week as you’ll start setting an expectation that it only takes them 10 minutes. When you then send them an hour of work, it might take longer than usual to get responses from the departments.

Ask for Approval for Both Ads & Individual Lines

Often in ad testing, you might test multiple calls to action for the same headline and description line 1. In these cases, you really do not need the full ad approved; you need just a line approved.

Explain the difference to the department to see if you can get some globally approved lines that you can use at will.

Now, a lot of legal departments won’t do this as they want to see the entire message and how the lines interact with each other before they will sign off on the ad approval. However, if you are going to do multi-ad group testing; then you will want both full ads and individual lines approved.

Notes for Rejected Ads

rejected adssNo matter how much you try to write ads that will always be approved, some will be rejected. The trick with rejected ads is to have a feedback loop as to why it was rejected. In some cases, it might be something as simple as an unsubstantiated absolute or a number in an ad. In other cases, it’s the combination of lines together that will cause the rejection.

Ensure that your ad review document has a place for notes on rejected ads. Many departments have a tendency to just reject ads without telling you why. This does not help out anyone. Explain to them that if you know why an ad is rejected, then you can create copywriter guidelines so that future ads will fall into the guidelines and that everyone will be more efficient in their work.

Now, you will have some ads rejected that are confusing to you as to why. In these cases, feel free to ask for an explanation. In some instances, you might even build in an appeals process for ads that you don’t think should have been rejected and you can explain why you think the ad is OK for approval.

This rejected ad feedback loop is crucial to improving your long term relationship with the department. When you create new ads, you do not want to have future ads rejected for using some line or claim that was previously rejected. If the department feels that you are ignoring their information, they will stop sending it to you. So while you can fight a little bit for certain rejected ads, in the end, legal and branding have more authority over your PPC message than you do.

Discuss Promotional Schedules

In some cases, your company will run promotions. These might be holidays, peak periods, or other marketing campaigns. Often, when big promotions are coming up, you need more ads approved than usual.

If you are mostly dealing with the branding department, then there is usually a big discussion on the promotion with the other channel teams. Make sure you are in this meeting so you can understand the promotion, the message, and the media to be used. In that meeting, discuss the ads you will need to get approved. The branding department might just give you a few lines to use for your current ads to match the message or you might need to create all the new ads for branding approval. As the promotion has a deadline (usually 1-3 weeks before the promotion goes live), agree on a date where you will give them all the new ads so they can approve them all with enough time for you to upload them all and make them active.

If you are dealing with legal, then you need to discuss promotional challenges with them. Explain why promotions are so important; and ask them if, before these big promotions, they can schedule more ad approval time than usual to make sure all your new ads will meet their approval. This is also another time to bring up individual ad line approval as they might just need to approve 2-3 lines to use in your current approved ads instead of them reviewing hundreds or thousands of new ads.

What To Do with All the Approved Ads

If everything lines up correctly, you will have more ads approved than you need to run at any one time.

In these cases, using features like draft ads will help considerably. With draft ads you can:

  • Write ads in advance and put them in a queue
  • Set conditions for when loser ads are automatically paused and the draft ad goes live

This will ensure that all the approved ads are in the system and that they will be used and tested.

image

If you can get individual lines approved, then you will often want to use Multi-Ad Group testing techniques to see which lines perform best for you across a large variety of ads.

When you regularly test ads, your account will improve. However, branding and legal departments are often considered barriers to proper ad testing. If you create a review process and feedback loop with the departments, they don’t have to be a barrier for ad testing and can often help you create an ad testing schedule so that your account continues to improve and grow its profits.

How Ad Rotation Affects Your AdWords Ad Testing

This post originally appeared on the adAlysis ad testing blog.

 
 
If you have multiple ads in an ad group; your ad rotation settings will determine how often each ad is displayed. Based upon how you are testing and your favorite metrics, you should consider the rotation setting you are using and how that affects your ability to receive statistical significance data to make testing decisions.

The Ad Rotation Settings

There are four ad rotation settings:

  • Optimize for clicks: the ad with the highest CTR should be displayed the most often
  • Optimize for conversions: the ad with the highest conversion rate should be displayed the most often
  • Rotate evenly, then optimize: the ads should have roughly equal impressions for 90 days, then the highest CTR ad will be displayed the most often
  • Rotate indefinitely: the ads should have roughly equal impressions

image

 

Ad Served Percentage

The ad served percentage shows you how often each ad was served across your account, campaign, or ad group.

image

When examining this data; it is important to keep in mind the time frame you are examining. If you have paused or deleted ads that were active during the timeframe you are examining, then your ad served percentages may not add up to 100% unless you show those ads.

In addition, it is useful to only examine the data when all the ads were running at the same time. If you create an ad one month ago; but you are looking at the last three months of data; of course it will look like the newer ad doesn’t have the appropriate ad served percentage; and it can’t as it wasn’t active for two of the three months you are examining.

Improper Ad Rotation

When you use any ad rotation setting except for ‘Rotate….’ (such as ‘Optimize for CTR’ or ‘Optimize for conversions’), sometimes AdWords makes decisions too quickly about which ad is likely to be a winner.

For instance, if the campaign had the ad rotation setting set to ‘Optimize for conversions’ and we examine a time frame where all the ads were active, we would expect to see that the highest ad percentage served would be the ad with the highest conversion rate.

image

 

Unfortunately, that is not always true. Sometimes AdWords makes decisions too quickly and the wrong ad has the highest ad served percentage. For example, the below campaign was set to ‘Optimize for conversions’. Ad two is the highest converting ad and has the highest click through rate; so by any ad rotation standards the second ad should have the highest percentage served. However, it does not.

 

image

This does not happen all the time by any means; however, it does happen. When your ads are served improperly, it affects your account’s goals and your ability to reach the minimum viable data in determining the true statistically significant winners.

 

How Ad Rotation Affects Minimum Viable Data

Any ad test should have a minimum amount of viable data, such as a minimum amount of time, clicks, impressions, and conversions. These may vary depending on the type of metrics you are using for ad testing and the type of keywords you are testing (such as brand vs product).

image

When your ad served percentages are skewed towards a single ad, then the other ads receive less impressions. Since they have less impressions, these other ads also receive less clicks and conversions. Since these ads are receiving less data, it takes longer for those ads to build up enough minimum viable data to make statistically significant decisions.

Use Rotate Indefinitely for Ad Testing

For most serious ad testers, you should be using ‘Rotate indefinitely’ as your ad rotation setting. This will ensure that your ads are equally served and that all the ads build up the minimum viable information to make ad testing decisions.

Even if you only care about conversion rate (and if you are in lead generation; there are probably better lead gen ad testing metrics) and think that using ‘Optimize for conversions’ is the best option; that is only true if you meet two conditions:

  • AdWords actually serves your highest conversion rate ad the most
  • You are removing losers as you have the data

This second point is an important one; take the metrics from this ad group that was not well looked after.

 

Ad % Served Conversion Rate Conversions
1 56% 9.0% 504
2 14% 8.3% 116
3 10% 7.2% 72
4 8% 6.1% 49
5 6% 3.2% 20
6 5% 1.1% 6
7 1% 0.9% 1
Total 100% 768

 

If ad 1 (with the highest conversion rate) had been served 100% of the time; then the ad group would have had 900 conversions instead of 768. As you should always be ad testing; there will be some opportunity cost when you test an ad that turns out to be a loser.

However, unless you are willing to test ads, you’ll never get better. Moreover, if your competition is diligent about testing, you’ll get worse by clinging to your old ads as they start to beat you.

Conclusion

To be a good ad tester, you should use ‘Rotate indefinitely’ as your ad rotation setting to ensure that all your ads are equally served. Then you need to be diligent about removing losers and creating more ads.

While this can be a lot of manual work, with good ad testing software, it can be very easy to find winner and losers and continuously test your ads so that your account keeps getting better and beats your competition.

The Best Ads Testing Metrics for Lead Generation Campaigns

This post originally appeared on the adAlysis ad testing blog.

There are many metrics you can test by; however, the best testing metrics for determining success vary by industry.

Often the lead generation industry is lumped together as a type of advertising campaign: just get me leads. It should not be classified this way. Within lead generation marketing campaigns, there are four main subsets of lead generation goals.

For the purposes of this article we will use CPA (Cost per acquisition) as the acronym. If you use CPL (cost per lead) internally, this would be the exact same metric.

The Types of Lead Generation Campaigns

I want the most leads regardless of incremental costs

These campaigns are commonly seen in long sales cycle, very high costs items, in the business to business world. There’s often little search volume; so all the competitors are fighting over the few leads that do exist.

There are exceptions where someone is just trying to lock-up the market or make a land grab where someone might overpay for leads for a while; but in reality, these campaigns are usually in high priced industries.

I want the most leads under a specific average CPA

In these campaigns, the blended CPA is what matters. It is OK if some leads are over the average and others are below it. However, the blended CPA needs to be within a specific target.

This is common in variable priced industries. You might have some leads that are worth $1,000 and others that are worth $10. It is usually difficult to determine any one keyword or ads value as the high value leads are often random and not predictable.

You might see this type of advertising goal in the ecommerce industry. For instance, I work with one ecommerce company that has some sales at $5,000 and others at $100. There’s no predictable way of knowing what ad or keyword will bring in the high priced sale – it’s completely random. Therefore, they work from an average CPA basis per sale instead of an ROAS basis.

I want the most leads possible, but no lead should be over $Y

For this marketing type, you work from a max CPA instead of a blended average. This is common when you are doing lead generation for other companies and then selling the leads. For instance, if you get a lead and then sell it to 4 others for $15 each, your breakeven CPA is $60. As you’d like to maintain margins, you don’t want any lead to ever cost more than $45, so you keep $15 per lead.

In this industry, if you get a lead that costs you $75 and you can only sell them for $60; you lose money on that lead. Since these are fixed costs, you have a max amount you want to pay for any one lead.

I want Y leads a month at the cheapest CPA

It is common for small businesses to not be able to scale properly. They don’t want the most leads if they can’t call them all back; therefore, they usually have a cap on how many leads they want a month.

The goal might be 150 leads a month at the cheapest CPA possible. In addition, these might have a cap such as get me up to 150 leads but don’t break $40 as an average CPA.

The Lead Generation Testing Metrics

There are several metrics that you can use for lead generation. Let’s first examine the metrics and then see which metrics best fit each lead gen effort.

CTR: Click through rate: Just because an ad has a great CTR does not mean it’s converting. This metric is not a good one to use for lead generation as it does not take conversions and cost into account. What this metric is good for is getting the most traffic regardless of how it converts.

CR: Conversion Rate: The conversion rate is a ratio between conversions and clicks. This metric does not take cost or volume (CTR) into account. What this metric is good for is ensuring that of the clicks you do get, you can tell what ad is most likely to get a conversion from those clicks.

CPA: Cost per action: This is your actual cost per action. CPA does not take into account volume (CPA) or the ratio of clicks to leads (Conversion Rate). This metric is good for ensuring that your leads do not break a certain cost.

CPI: Conversion per Impression: Conversion per Impression takes into account volume (CTR) and conversion rate. This is a great metric for determining which ad will result in the most conversions possible. However, this metric does not take into account cost.

Testing by More Than One Metric

You do not have to use only one metric to choose winners. You might want to only choose winning ads based upon an ad winning multiple metrics. Now, what you might find in some cases is that one ad will win by one metric and another by another metric. In these cases, you can pick an ad by your top metric, and then create a new ad looking to find one that will be a winner in multiple metrics.

For example, in this ad test one ad has the lowest CPA and another ad has the highest conversion per impression:

image

 

Therefore, if you wanted the most leads regardless of cost, the second ad with the highest Conv. Impression (CPI) would be your winner.

If you had an absolute lead cost; such as £7; then the first ad would be your winner since the second ad is above your lead cost; however, your target lead cost was £15; then the second ad would be your winner as it will provide the most leads and it is below your lead cost.

In some marketing campaigns, you will only use one metric to determine winners. In others you will use more than one metric. Therefore, let’s examine which are the best lead generation testing metric for each type of lead gen campaign.

The Best Metrics by Lead Generation Type

I want the most leads regardless of incremental costs

In this lead generation type; the goal is absolute leads regardless of cost. Therefore, you can use a single metric Conversion per Impression to determine your ad winners. As this metric takes into account volume (CTR) and conversion rate (CR) it will be your testing metric of choice.

In this case, ad 2 in the winner as it has the best CPI even though it is not the CPA (Cost/Converted click) winner.

image

I want the most leads under a specific average CPA

These campaigns care about volume and average CPA. Therefore, we will want to use two metrics to determine winners: CPI (volume) and CPA (cost). When you find an ad that is a winner in both metrics, then you have found a great winning ad. However, you have possible combinations. Because CPI speaks to volume and the goal is the most leads (within a cost); we’re going to pick ads based upon the winning CPI ad (and the other ad/s will be the CPI losers).

  • CPI winner. CPA winner. Fantastic, this is your best ad. Feel free to test against it; but you have a top notch ad.
  • CPI winner. CPA loser; but the CPA is below your average. In this case, your top CPI ad is your winner as it has the most leads below your lead cost.
  • CPI winner. CPA loser and CPA is above your average. The other ad has a CPA below your target.
    • In this case, the ad below your target CPA will be your winner as ultimately you will be judged on how much a lead costs.
  • CPI winner. CPA loser. All ads are above your target CPA.
    • In this case, you have a choice: what matters more – cost or volume?
      • If its cost, you will pick the lowest CPA ad, examine why the other ad did better on a CPI basis, and then create a new ad to test
      • If its volume, you will pick the highest CPI ad, examine why the other ad prequalified people better (common for lower CPA), and create a new ad to test

In this case, there is a winner by CPI (conv. / impr) and CPA (Cost/Converted click) and thus that is a clear winning ad for this lead gen type.

image

I want the most leads possible, but no lead should be over $Y

The management of this type of account (as most should be) should start at the keyword and search query level to make sure you are only using keywords that are under your target CPA. From an ad testing standpoint, we should use cost (CPA) as a primary metric since the main goal is not to break a certain CPA and then use volume (CPI) as the secondary indicator. Since CPA is the prime metric, we will examine the ads based upon the CPA winner (and the other ad/s will be the CPA Losers).

  • CPA winner (and CPA below your target). CPI winner. This is your winning ad. Feel free to test new variations; but this is a clean win.
  • CPA winner (and CPA is above target). CPI winner. This is your winning ad since it’s the lowest CPA with the best CPI; but since the CPA target is above your goal, you will need to be aggressive about testing new ads.
  • CPA winner. CPI loser. In this case, we need to establish what you care more about: the most leads or the lowest lead cost
    • Lowest lead cost: In this case, you will take the CPA winner/CPI loser ad as that will be the lowest lead cost. Examine the CPI winner, pause it, and create a new ad to see if you can find a combination CPI/CPA winner.
    • Most Leads: CPI winner has a CPA below your max CPA.
      • In this case, your CPI winner will be the winning ad and not your lowest CPA ad as you want the most leads and both ads are below the target CPA. Again, examine the losing ad, pause it, and write a new ad that can become a CPA & CPI winner.
    • Most leads: CPI winner has a CPA above your max CPA.
      • In this case, your CPA winner will be your winning ad since your top CPI ad breaks the primary marketing goal of an absolute max CPA. As usual, examine loser, pause winner, create new ad.

If our target CPA was £10 and our goal was the most leads; then the 2nd ad would be our winner as it has the higher CPI (conv. impr) and is still below the target cost.

If our target CPA was £9; then the first ad would be the winner as the 2nd is is below our lead cost.

If our target was £10 but we cared more about cost than total leads; then again the first ad would be our winner as it has the lowest CPA.

 

image

I want Y leads a month at the cheapest CPA

For this campaign, your primary goal is a set number of leads each month. While the account states you want them at the cheapest CPA; there is usually a max CPA cap as well; however, the main goal is cheap CPAs hitting a specific conversion target. Therefore, your top metrics will be CPA (you want the cheapest leads) and conversion rate (making sure the clicks you do get actually convert).

While CPI could be substituted for conversion rate to find the most volume (up to the point where you can pause the account as you hit your goal) at the cheapest CPA, these accounts are usually small businesses with few clicks, so it usually takes a long time to get CPI confidence; and therefore, it’s usually easier to use conversion rate as the secondary metric. If you have a high number of target leads, then you can substitute CPI for CR (conversion rate) in this evaluation.

In addition, your traffic is split between two or more ads. So when you examine the timeframe (say previous month) you would multiply the number of conversions by the number of ads in an ad group and then take into account the percentage of the month being used to examine the data to get extrapolated conversions. This would show you what would have happened if only the winning ad were running.

For example, if your goal is 100 leads a month and you have three ads in the ad group and you examine a week of data. Then you would take the number of leads that ad received (let’s just say 10 for example purposes) x 3 (number of ads in the ad group since the ad was sharing impressions) x 4 (4 weeks in a month, you only ran the ad 1/4 of the time, you could also divide this by the percentage of time; such as 25%). That means that our extrapolated conversions for the ad would be: 10 (conversion) * 3 (ads) * 4 (weeks) = 120. That means if our winning ad was the only one in the ad group; the ad group would have hit the max conversions.

Now as there is usually more than one ad group in a campaign, odds are that you can’t do this across all ad groups easily. Hence by just using CPA and conversion rate we can usually arrive at the same results for the small amount of data the ad groups usually have.

  • CPA winner. Conversion rate winner. Fantastic, this is your winning ad. Feel free to test, but you have an absolute winner.
  • CPA winner. Conversion rate loser.
    • In these cases, you need to look at the overall account. Are you close to your target leads?
      • If yes, your CPA winner is likely to be your winner since a few more leads per month, even at a low CPA might push you over the edge for your account lead goals
      • If no, if your conversion rate winner’s CPA is somewhat close to the losing CPA (so you aren’t going to suddenly spike your CPA)?
        • If the CPAs aren’t too far apart then your highest converting ad will be your winner since you need more leads
        • If the CPAs are dramatically different, then in most cases you will pick the lowest CPA ad as picking the highest converting ad could dramatically increase your overall CPAs.

As usual, when you find losers, pause them, learn from them, and create new ads that will help you reach your goals.

In this case, ad 1 is going to be our winner as it has the highest conversion rate and the lowest CPA.

image

 

The Quality Score Wrinkle

The one issue we should cover is how quality score affects lead costs.

Google does not show any quality score information at the ad level, it is all at the keyword level. However, when examining ads, usually the higher the CTR, the higher the Quality Score for that ad and keyword combination. This can mean that if you have an ad with a much higher CTR than the other ads in the ad group; it could have a higher quality score and thus you pay less per click (or it has a higher position).

This is why you can have one ad that has a higher conversion rate than another ad, but also a higher CPA. If both ads had the exact same CPC and exact same conversion rate; their CPAs would be the same. However, if one ad has a higher quality score, and thus a lower CPC, it can have a worse conversion rate and yet a better CPA than another ad.

Ad Quality Score CPC Conversion Rate CPA
1 high $0.75 5% $15
2 low $1 6% $16.67

 

Therefore, even if you are not picking winning ads by CTR: it is useful to examine CTR information when creating new ads.

Easily Automate Your Ad Tests

The next time you create a lead gen account; don’t just think its yet another lead generation account – determine what type of lead gen account it is. By understanding the type of lead gen account; you can then determine how to create and manage your ad tests.

Now, determining statistical significance by all of these metrics can be a tremendous amount of work. If you are looking for a solution to automate the analysis of your metrics as well as only seeing alerts when you have winners, take a look at AdAlysis: Powerful Ad Testing made Simple.