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


Join us for an In-Depth Webinar on Quality Score


We’re joining the Acquisio team for an in-depth look at Quality Score that’s open to anyone on June 2nd.

There’s a few topics we’ll cover in the webinar.

Quality Score Importance

Quality Score is one of the most important and least important numbers in your AdWords account at the same time.

  • When its bad, it is a very important number.
  • When its good, you can ignore it for the most part and only focus on specific areas where it can be improved.

It can be difficult to find out if its good or bad or where the areas of opportunity lie. We’ll go over how to uncover areas of opportunity.

Quality Score Improvement

There are a lot of myths about quality score from how much you actually save in increasing your quality score to what the factors really are that matter.

In the webinar we’ll go through some of the ‘price discounting’ myths surrounding Quality Score and show you what actually happens as your quality score goes up and down.

Quality Score Factors

Quality score is made up of a few factors, and each one seems a bit convoluted (what is relevance?). We’ll go through the factors, their importance, and what they really mean.

New Data Insights

AdAlysis recently added many quality score graphs and data which has lead us to uncover some new data that we’ll reveal in the webinar.

Join Us

If you’d like to learn more about quality score, how its created, and how to improve it, please join us for the webinar.

Join us for PPC Rockstars Tomorrow


PPC Rockstars is a great weekly podcast by David Szetela on anything and everything PPC related.

Tomorrow, Brad Geddes is going to be a guest on PPC Rockstars and chat about:

  • Quality Score (reverse engineering the various factors)
  • The right hand side rail changes
  • The death of phrase match and the rise of broad match
  • And a few other things in-between

The show airs at 1 pm ET on April 7, 2016; and will be available as a download (or on iTunes, Google Play, etc) in the next few day from PPC Rockstars.

Should you bid on non-branded keywords that you already rank organically for?

One of the most common questions in paid search, is should you bid on keywords that you already rank well for organically, or is it just a waste of money?

In this article I’m going to be examining the two leading case studies by Kenshoo and Brad Geddes to provide a data backed answer to this question. I’m also going to be identifying other reasons why it may be worth bidding on organic terms such as the ability to choose your own landing pages, to increase your conversion rate and being able to use your PPC data to optimize your organic search ads.

1) Paid search allows you to increase your overall profit (in the vast majority of cases)

Case study 1 – Kenshoo and HP –

A recent joint Study by Kenshoo and Resolution media into Hewlet Packard, that looked at 1 year worth of data and $10 million in direct revenue found the following results. The test only featured data where there was a paid listing and organic listing for the same keyword.

When there was a number 1 organic ranking users still clicked on the paid listing 39.4% of the time and Paid search revenue accounted for 41.6% of revenue where there was a number 1 organic listing too.



The study also found that paid search drove 3.2% more revenue per visit than organic search traffic.

Case study 2 – Certified Knowledge

Brad Geddes ran a test with one of his clients, over various different timeframes and for a range of different keywords and got the following results.


Keep on reading!

The Best Hidden Features of the AdWords Editor

Along with a browser and Excel, the AdWords Editor is probably among the most used software for any serious PPC specialist.

In December 2014 Google released the biggest update to the Editor since its release back in 2006. They redesigned the Editor from the ground up and included many new features – some prominent and others more hidden.

In this post, I will show you some of the most useful but lesser known features of the AdWords Editor.

1. Shortcodes for smarter bulk edits

One of the best hidden gems in the AdWords Editor is the shortcodes. The AdWords Editor supports a list of shortcodes, which can speed up your work when combined with the Editor’s build in replace function. Google actually don’t call them shortcodes but “formula words“.

Available shortcodes in the AdWords Editor include:

  • [account] – Account name
  • [campaign] – Campaign name
  • [adgroup] – Ad group name
  • [keyword] – Keyword text
  • [headline] – Ad headline
  • [description1] – Ad description line 1
  • [description2] – Ad description line 2
  • [displayUrl] – Display URL text
  • [finalURL] – Final URL text
  • [field] – The original text of the field you’re changing

So how exactly is this useful you may ask.

One use for it could be to automatically insert the campaign or ad group name into your ad text. Suppose you have 10, 50, or even hundreds of campaigns each named after the different geographic locations they cater to. In each of these campaigns you have an ad group for “cheap hotel” searches. And because you were in a hurry you just pasted the same general ad in each of these ad groups. Not exactly the best ad text for local searches.

Luckily, with the shortcode [campaign] you can now use the replace function to make every ad more specific. If the headline of all ads are “Cheap hotel” you can use the replace function (Ctrl+H will bring it up quickly) to find the text “Cheap hotel” and replace it with “Cheap hotel in [campaign]”.

Notice how you can choose to only perform this action in the ad headline. This way your ads will now have been changed to reflect the name of the campaign.

As you can imagine, there are many ways to use these shortcodes. To get you started you can find more examples on the official help page.

2. Hiding empty types from the navigation

One of the things I hear most people complain about in the “new” Editor is the completely changed navigation. In her post Why The New Adwords Editor Sucks, Mellissa Mackey wrote the following about the new navigation:

Not only has it moved to a totally different location on the page, but it’s now in text only, rather than a picture-like graphic layout. It’s easier to find what you’re looking for in a graphic square, like the old Editor, rather than in a long list of text, like the new one.

And I agree – it is harder to quickly find what you are looking for. But fortunately, Google included a nice little feature called “Hide empty types” which helps keeping the type list tidy.

To access this feature you will need to click the drop-down menu button to the right of the “Manage” header. Then you can click “Hide empty types” in the drop-down menu, which will hide all types that don’t currently match any of your ads.

Just remember to deactivate it when you need access to types that are currently empty.

If you find it to be too much of a hassle to turn this setting on and off all the time, you can actually choose the specific types you want to see under “Custom”. This way, you can hide types you only use rarely.

While this feature won’t bring back the old navigation it will make the new navigation a bit more tolerable.

3. ‘Flexible reach’ for setting the targeting options for RLSA

One of the AdWords mistakes keeping me up at night is choosing ‘Target and Bid’ by mistake when setting up RLSA. Doing this will drastically reduce the traffic as the audience is narrowed to only returning visitors, buyers, or whatever audience chosen.

In the AdWords Editor you can copy the audiences from one group and paste it to all your groups. But even though you copy the audiences you don’t copy the targeting setting, and sometimes you’ll find some of your ad groups set to “Target and bid” while the rest are set to “Bid only”.

Therefore, I always recommend you check the “Audiences” area in the editor to make sure everything is set to the correct targeting option.

Changing the targeting option from “Target and bid” to “Bid only” (or the other way) is rather hidden in the Editor. To find it you need to go to “Ad groups” and choose the “Flexible reach” tab as shown below. Here, you can select every ad group and then make sure either “Bid only” or “Target and Bid” is selected under “Interests and remarketing”.

4. Duplicate ads or keywords within their respective ad groups

Let’s say you have a number of different ad groups each containing a single tailored ad and now you want to create a second ad inside every ad group in order to start split testing different messages.

So how would you duplicate the ad inside every ad group? Keep in mind the ads in all the ad groups are not identical, and therefore it’s not an option to just copy one ad and paste it inside every group.

Sure, one option is to manually go through every ad group and copy/paste the current ad, but this is neither fun or good use of your time.

The smart way is to select every ad across all your ad groups and bring up the replace window (Ctrl+H). Here, you can make your change and then you must tick the “Make changes in duplicate items” field in the lower left corner.

By doing this you will make a modified duplicated ad inside every ad group.

5. Exporting account to HTML

Many were disappointed to see that the HTML export feature was removed from version 11 of the AdWords Editor. This feature was extremely useful in communicating account setup to clients as this would give them a preview of the structure including ads and keywords.

But in November 2015, Google released an update to the new editor bringing back the HTML export feature.

To export an HTML summary, you need to select “Account” in the menu, then “Export”, and then choose whether you want to export the whole account or just the selected campaigns and ad groups.

Name the file and make sure to choose “HTML files” as your filetype. This will give you a nice looking browsable HTML document as the one shown below.

The different elements are even clickable and will expand to reveal more information like targeting or ad groups.

You can now share this file with anyone you want, regardless of whether or not they have access to the AdWords Editor.

This is a guest post by Frederik Hyldig, Head of PPC at s360 – one of the leading search agencies in Denmark. Frederik has been featured on PPC Hero, Wordstream, Moz and other international online marketing blogs. Follow him on Twitter or his personal blog.

Learn How To Test Mobile & Desktop Ads Correctly

Last year, Mobile search surpassed desktop search in terms of total queries. The statements of  ‘next year is the year of the mobile device’ have passed. Mobile has arrived.

However, most account we view still treat mobile and desktop users the exact same. If there’s a variance, its because of a mobile landing page – not because the account owner has thought about how mobile and desktop users behave and what they wish to accomplish by device.

As mobile and desktop users are often looking to accomplish different tasks based upon the device type; you should test ads differently based upon the user’s device.

However, first we should examine the concept of ‘effective device’ so you understand where your ads are really showing.

Examining Effective Device

With Google and Bing, you can designate an ad as ‘mobile preferred’.


However, that ad might not only be on mobile devices based upon the other ads in an ad group. For instance, if you only have mobile preferred ads in an ad group; then those ads are displayed on desktops and tablets.

If your ad group only contains desktop ads and your campaign is showing on all devices, then your desktop ads are effectively ‘all devices’. This leads us to an important concept: effective device for ads.

There are three possibilities for effective device:

  • All devices: These ads are showing on all devices
  • Mobile: These ads are mostly showing on mobile devices
  • Desktop: These ads are mostly showing on desktops and tablets

I do use the word ‘mostly showing’ on purpose as it’s possible due to various technology reasons that your mobile ad might show on a desktop or vice versa. This doesn’t happen often, but it does on occasion.

Therefore, to control ad serving by device you need at least one desktop and one mobile device ad in an ad group. To test by device, then you need at least two desktop and two mobile preferred ads in an ad group.

Determining Your Effective Devices

There is not an easy way in AdWords to determine what devices your ads are running on nor to determine if you are testing your ads by device.

The easiest way to get started in understanding this information is to follow these steps:

  • Navigate to the ads tab in your account
  • ‘Customize the columns’ and ensure that device preferred is selected
  • Download the data
  • Create a pivot tablet from the data
  • In your Row Labels; select ad group (if you have a large account, you can select both campaign and ad group)
  • In your Columns, choose device preference
  • Finally, your values should be counts of ads

Now, that will just give you a bunch of numbers that are hard to read; so to make the pivot table visual, add conditional formatting to the pivot table:

  • If column is 0, then color it red
  • If column is 1, then color it yellow

The legend of effective device:

  • If the ‘all’ device column is red (so there are no desktop ads); then your mobile ads are showing on desktops; meaning your mobile ads have an effective device of ‘all’.
  • If the mobile column is red, that means you don’t have mobile ads in the ad group and your desktop ads are showing on mobile devices

The legend for testing:

  • If any column is yellow, then then you have an ad of that type, but you aren’t testing in that ad group



Where to Start Creating Ads

There are two places where you will want to sort to determine where to create ads:

  • Cost: test where you are currently spending the most money
  • Impressions: test where you have the most opportunity

The easiest way to do these sorts is to:

  • Add spend and cost to the pivot table
  • Copy the pivot table data (with formatting) to a new sheet
  • Filter by the ‘all device’ column and only include cells that are highlighted red (your mobile ads are showing on desktops)
  • Sort by spend or cost highest to lowest

That will ensure that you have desktop ads in all your ad groups. Now repeat the process filtering by the mobile column by ‘red’ or zero mobile ads.

Once you have at least one ad by device type by ad group; then you are controlling ad serving by device.

Next, you will want to repeat these steps for the columns that are yellow so you can make sure you are testing your top ad groups for each device.

Automating the Process

This process will give you a snapshot of your ad and testing information in the account at the moment you created and analyzed the pivot table. However, it won’t give you this information automatically until you create the table and analysis again.  In addition, just because you are testing doesn’t mean you are declaring winners.

If you want to easily:

  • Receive alerts by device type
  • Have a simple way of sorting this data by cost or impressions (or other metrics)
  • See when you have ‘winning ads’ by effective device

Then please take a look at which will make this processes simple and automated.


To be a true PPC expert, you need to move past how to just create and test ads across all devices. You need to think about how users interact with your ads on each device and across devices. It’s useful to initially lay out some hypothesis about users and do multi-ad group testing. Then once you’ve learned at a high level how users interact with your ads, move towards testing your ads by effective device. If you want to see some lessons on how to do this; take our 7 day free trial here at Certified Knowledge and watch the videos on ad testing.

Is Phrase Match Dead?

As modified broad match continues its march to take over being the absolute best match type available; many accounts have ceased to use phrase match. While we still hear people spout the benefits of exact match, the match type that is in the largest decline is phrase match.

In many ways, this match type isn’t bad – it’s misunderstood. Let’s compare modified broad to phrase match.

Can show for: Phrase Modified Broad
Misspellings Yes Yes
Stemmings, word variations Yes Yes
Keywords can be in any order in the query No Yes
Keywords must be in the same order as the query Yes No

The biggest difference between these match types is the word order requirements to show.

With phrase match, the keyword must be in the same order in the query as they appear in your account; but the words can show for misspellings, plurals, singulars, and so forth.

With modified broad match, the keywords can be in any order in the query and can show for plurals, misspellings, and so forth.

So what does word order really mean to search queries?

Let’s take a look at the match type data from two accounts.

Account 1

Match Type Conv. Rate CPA
Exact 3.1% $56
Phrase 2.9% $61
Modified Broad 2.9% $61


Account 2

Match Type Conv. Rate CPA
Exact 4.5% $27
Phrase 4.4% $29
Modified Broad 1.2% $74


There is a huge difference in the CPA variance and conversion rates from phrase to modified broad in these two accounts. In account 1, there’s no change from modified broad to phrase match. In account 2, there’s a huge difference in the data between these two match types.

What Does Word Order Really Mean?

The first account is a financial lead generation company, so these queries are useful to them:

  • Chicago mortgage
  • Mortgages in Chicago
  • Chicago home financing
  • Home financing Chicago

As you can see, switching the word order around doesn’t change the user intent. The searcher is looking to get financing in Chicago for a home.

Here’s the two most popular modified broad match queries for account 2:

  • Contractor license
  • Licensed contractor

Google doesn’t make a distinction between ‘license’ and ‘licensed’ with the changes that occurred to variation match; so these are essentially the same words. However, their order means a lot.

A Contractor License is license that a person needs to acquire to be considered a licensed contractor. This search is conducted by someone who is looking to obtain a license.

A Licensed Contractor is someone who has already obtained their license; and this search is conducted by someone who is looking to hire a contractor with a license.

The word order is so important to this account (it helps a person to obtain their license) that modified broad rarely does well. They can use lots of negative phrase matched keywords; but in the end – the simple word order change is the difference between wasting money and obtaining a good lead.

What if Word Order Doesn’t Matter – Should I Use It?

Let’s consider these stats for one account:

Match Type Conv Rate ROAS Profit/Sale
Exact 2.9% 463% $57
Phrase 2.4% 405% $51
Modified Broad 2.3% 395% $49

In this case, modified broad and phrase match are fairly close in profit/sale, ROAS, and conversion rate.

There is one number missing – total sales.

If this company had 100 sales/month from phrase match vs modified broad match – it would make an addition $200 per month. If the company had 10,000 sales a month, it could make $20,000 more per month by splitting out the match types (this is back-of-the envelope match without splitting up all the sales by match type).

Managing all 3 match types, assuming phrase and modified broad are close in their metrics, is more work; but it is often more money. The question is: how much more money?

At 10,000 sales/month – it is worth it to pay someone or use a 3rd party bid system to manage the additional profit.

At 100 sales/month – it is usually not worth it to do all the extra work of managing the additional match type (and all the additional keywords and ads that might go along with the additional match type).

Phrase Match Isn’t Dead – It’s Just Misunderstood

Phrase match has a lot of uses. It’s very useful when you are managing a lot of sales and the sheer number of your sales means you can eek out more profit if you manage the bids by match types correctly.

Do most small local business accounts need to use phrase match – no.

Do some of them – absolutely.

Phrase match is incredibly useful when the word order matters. Word ordering doesn’t always matter; but when it does – phrase match is an indispensable match type to employ to make sure you are reaching the correct searchers.

The Most Common PPC Mistake that is Never Mentioned

The absolute most common PPC mistake is actually rarely talked about. It never makes the top 10 mistake lists; and it’s usually pointed out by someone else – in this case via Twitter.

A special thanks to Robert Brady & Sam Owen in reminding me to check these mistakes.

If you use one of the editors (AdWords or Bing Ads) to import keywords, mass create ads, etc – you always have a header row in Excel to remind you of the columns:


However, when you copy and paste your huge tables into the editors, most people hit ‘Cntl-A’ (select all) from Excel and then ‘Cntl-V’ (paste) into the editor.

When you upload those files, you get ads like this:


Another common mistake is in the AdWords interface itself. When you add more keywords, the default text is: Add Your Keywords Here

If you don’t remove that; your ads suddenly show up for a variety of search queries, and since its broad matched by default; some of the matches can be quite amusing.


Here’s the secret in the bulk import. Instead of copying and pasting all, in Excel do this:

  • Highlight the second row (past the header)
  • Then press & hold: Cntl – Shift – <down arrow or down>
  • Then hit Cntl-C for copy
  • Paste into the editor as usual
  • Upload

Of course, you can also just delete the header row and then copy and paste.

This is also why I’m a fan of filtering my keywords and ads by ‘keyword’ or ‘placeholder’(or whatever you use that needs to be replaced in your ads)  on occasion just to make sure I’m not spending money on irrelevant keywords & ads.

7 Great Uses for Broad Match (yes, really! broad match)

Broad match keywords have a bad name, and rightfully so by performance marketers; however, there are times that using broad match is useful.

Please note, we’re not talking about modified broad match, which is often very useful to use – only broad match. If you’re wondering why we consider broad match so terrible, just consider these examples.

Keyword Actual Search Query
Wedding cakes How to make a Dora cupcake
Flower delivery McDonald’s Delivery
Nike Tennis Shoes Restaurant waiter clogs
Electric company iPod charger


In each of these cases, the demographics and intent of the user’s query was completely different than the keyword’s idea; yet these keywords triggered ads for these queries.

While using broad match is often considered one of the biggest mistakes any account can make; it has several good uses.

1. Research

Sometimes you want to understand the entire universe of semi-related keywords. Broad match is useful for finding all of these relationships. Just consider that it should be a ‘research budget’ and it should not count against the PPC teams’ ROAS or CPA goals.

Note: If you want to understand all the related queries for your site, not just words, we recommend using DSAs (dynamic search ads) to accomplish this goal.

2. Understanding a Change in Query Volume

There are industries that have a lot of potential risk in them. For instance, a car recall can put a large strain on a company. Often these companies want to buy some broad match terms to get ahead of possible product problems and determine if there is something faulty can it be easily fixed, do they have what’s necessary in place to deal with the PR that might arise, and so forth. Buying broad match in these types of industries is a legal and PR measure to understand what might happen and to be ahead of the curve.

3. Advertising to Very Small Geographies

If you’ve been in PPC long enough, you’ve been asked to sell a niche product to a city that’s so small that every relevant word is ‘low search volume’ and there’s no way to advertise to the region effectively. In these cases, you often have to use some broad match just to get enough query volume to make your keywords display. The advantage here is that the competition is usually so light that the bids are very low; so even though you get stuck with a lot of irrelevant queries; you can still maintain acceptable returns for the company’s marketing budget.

4. Multi-Lingual Countries

This is one of my favorite broad match uses – using English words to target non-English languages.

Did you know that broad match can show for the same query in a different language? None of the other match types will trigger ads if the query was done in a different language – broad match is an exception.

In some countries, many search queries are in multiple language, but the country’s inhabitants speak and read English (I’m using English as an example, in many countries this could be French, German, Arabic, etc).

In this account’s target country, there is a lot of Arabic search volume, but everyone speaks and reads English. The ads and landing pages are all in English; and by using English broad matched keywords; we can capture some Arabic search volume and conversions without having to support an entire Arabic account and website.


5. When Queries Cross Languages

What happens when the query is in two languages? I see a lot of queries that contain English and Arabic, English and German, English and French, etc. If you thought years ago trying to manage all the possible misspellings of keywords was difficult (you no longer need to do this – once upon a time you did) – try mixing and matching keywords that comprise multiple languages.

In these cases, broad match is very useful.


6. For Non-Latin Based Languages

As a general rule, the further a languages roots are away from Latin – the better broad match is to try as it often performs closer to modified broad match than what we think of as broad match in English.

Broad match in Spanish performs fairly similar to broad match in English. However, broad match in Arabic, Cyrillic, and many other languages can do quite well.

7. Capturing the Uniqueness of the Long Tail

Modified broad match will capture a lot of the long tail; but not all of it. As voice search changes how people are searching, users are conducting queries such as:

  • What song is this? (i.e. holding up a phone and listening to the radio)
  • Who sings this?
  • Where can I buy it?

In that example, and artist name was never used. A song name was never used. The entire query chain is contextual. As broad match can show for ‘session based matches’; where an ad is shown based upon a previous query (and Google knows the context) some of these new phone interactions aren’t possible to capture with your traditional keywords.

As 15% of search queries have never been done, or haven’t been done in at least 3-6 months; you need broad match to capture the entire long tail. Now, don’t start by just thinking you need all broad match; you must manage negatives with broad match and add queries as they do convert. However, if you want to capture everything (and that includes the bad with the good); then you need to use some broad match.


Broad match in and of itself is not a bad match type.

It’s usage and the expected results are often to blame for how poorly it performs in many accounts.

As a general rule, there are two times to try broad match:

  1. You’re willing to pay to learn something about your keywords or to reach everyone
  2. You’re advertising in languages that are not Latin based

Do you have any favorite broad match uses or horror stories?

Join Us for a Look at 2016 PPC Predictions

It’s 2016: Are you thinking about what the year is going to bring to PPC?

I’m doing a webinar on Thursday (January 28) with two of the industry’s best about what’s coming to PPC, some big recent changes, and we’re going to look into the future of paid search.


You will hear from Marc Poirier, the CEO & Co-Founder of Acquisio. Fred Vallaeys, the CEO of Optmyzr and long time industry veteran, and myself (Brad Geddes), a 17 year vet in PPC.

We’re going to hit a lot of topics on a lot of topics in a nice roundtable discussion in a short amount of time. We plan on touching on:

  • The changing dynamics of agencies
    • Should you even target enterprise companies anymore?
    • What does marketing automation and customer journey marketing mean to you?
  • The changing world of targeting
    • How audience targeting and customer match works
    • Is this going to compliment or supplant keywords?
  • What is going on with Bing/Yahoo?
    • Verizon owns AOL
    • AOL is now powered by Bing
    • What’s up with Yahoo?
  • Is voice search all it’s cracked up to be?

We’ll take audience questions and look into our crystal balls and try to predict (or at least trend) what’s on store for the future of PPC.

Registration is free, but the seats are limited. You can signup for the webinar here: 2016 Predictions Webinar.