Brad Geddes / PPC Geek
Official Google Ads Seminar Leader.
Author of Advanced Google AdWords.
Co-Founder, Adalysis.
(703) 828-5811‬
Brad Geddes's Theories on Marketing When Best Practices Fail

When Best Practices Fail


Best practices are called such because in most cases they work. In the PPC world, they are a set of processes or layouts for which conventional wisdom has concluded they are the best method for creating or managing accounts and landing pages.

Conventional wisdom is an idea that is widely accepted because it is usually true.

Just because it is usually true does not mean its correct for you. You should question conventional wisdom to see if it holds true for your situation.

In today’s article, I will examine some cases where conventional wisdom was inaccurate so you can start to judge your own accounts outside of the standards of conventional wisdom.

Never Force Account Creation Before Shopping Cart Checkout

In the vast majority of cases, forcing account creation lowers conversion rates. The goal for an ecommerce site is to put as few barriers as possible between viewing a product and processing a credit card. I’ve seen companies increase revenue as much as 40% by just removing the account creation process from before checkout to after checkout.

Then there was the time that decision was unwise.

I was working with an etailer and when we removed the forced account creation; the conversion rate did go up as expected. However, the lifetime revenue of the customer declined. When a user had an account and could save credit card, address, wishlists and other information within their account, users were more likely to buy a second and third time on the site.

After doing some testing and a lot of math, we realized that the total revenue was higher when users were forced to create account even though the initial conversion rates were lower.

Forced account creation was put into place on purpose.

Always Put Calls to Action Above the Fold

There are numerous studies showing how the most valuable real estate on a page is above the fold. That just by changing the call to action or moving a checkout button to the top of the page can have a dramatic effect on conversion rates.

And then there are the exceptions.

For a lead generation site we had a nice tight form with benefit statements alongside and it was doing quite well. The design team had created lots of media for some offline promotions and really wanted to showcase some of their design on the landing pages. After some debate, the media was added to a test page which forced the conversion actions to below the fold.

The conversion rates went up.

Next, the media above the fold was shrunk to move the calls to action above the fold. The conversion rates went down. After over a dozen rounds of testing, a very clear pattern emerged: conversion rates were higher when the call to action was below the fold.

No one could believe the findings. So another dozen rounds of tests were implemented. The results did not change. I haven’t seen this often; but I have seen a few cases where conversion rates are better when the main call to action is below the fold.

Never Send All Traffic to the Homepage

This is usually PPC 101. Find the page that answers the searcher’s question and send the traffic there. If you can show geography or industry (or both) information on the landing page the conversion rates will be even higher.

I have seen this overall concept fail a few times.

The first time was for a site buying a lot of paid search. They had a geographic aspect so we created lots of landing pages that showed the city skyline, actually product examples from that city, etc on the landing pages.

They did wonderful for the display network.

They totally failed for the search network.

After many tests, the company could not beat their homepage for search. All their search traffic goes to the homepage. All their display traffic goes to pages created for an easy conversion funnel.

This concept is sometimes echoed in lead gen sites where even though there are queries (such as Chicago insurance) the landing page chooses to ignore the geographic data and asks a simple ‘zip code’ question on the landing page. Often this works because it gives someone a very easy first step to try and get them invested in the process of moving through the form fills on a site.



Other times that does not work and geographic pages work best. In either case, test them out for yourself.

Broad Match Will Not Convert Better Than Exact Match

So technically this is always a true statement. However, in reality it does not always work so smoothly.

If you have several low volume terms that you add as exact match and then receive the warning ‘these terms are not being displayed due to low search volume’; then often you need to keep a phrase or modified broad match of a similar keyword in the account to catch these terms.

I was working with a medical company where the misspellings outnumber the correct spelling for many of their keywords. Few of the misspellings had enough volume to be displayed. Google was not matching the modified broad match far enough to capture all of the misspellings. Its amazing how many ways even common words can be spelled, let alone medical jargon terms.

Therefore, the only way to capture all of the misspellings was to include broad match terms. The misspellings had higher conversion rates than the proper spellings. In the end, the broad match variation of the word had a higher conversion rate than the exact match.

The broad match is now in its own ad group with a plethora of negative keywords, including the negative exact match, yet with a higher CPC than the exact match version of the same word.


There are many more examples of best practices failing:

  • Pages with no calls to action performing better than pages with calls to action
  • Ads for expired holidays outperforming other ad tests (yes, there are still Valentine’s day and Black Friday ads running on purpose)
  • Autoplay video working in a B2B environment
  • 15 minute YouTube videos with lower CPAs than 3-5 minute videos
  • And the list goes on…

When you are first starting, following best practices or conventional wisdom is a smart move: Learn from the mistakes of others.

However, once you understand not only what the best practices are, but why they are best practices – then you can test and judge these assumptions for your own account.

It’s only through testing everything for yourself that you can move beyond the conventional wisdom of others to relying on your own set of best practices.

No Comments

  1. lisasanner
    June 26, 2012 at 10:26 am · Reply

    Great article Brad! So true. Providing solid explanations/reasoning when/why best practices fail is one of the hardest parts on my job but always keeps it interesting and fun. Nice to know I’m not alone. 🙂

  2. rocio97
    June 29, 2012 at 6:43 pm · Reply

    This was all great advice! The hardest thing for me is finding non competitive exact matches. Thanks for the post.

  3. pkilchermann
    July 11, 2012 at 2:59 pm · Reply

    Indeed – a thousand times, indeed.

    It could very well be that I know little about adwords, but I have a lot of wild stuff happening: my broad match has twice the CTR and Conv. Rate as my exact match, for my most popular keyword, for example.

Leave a Reply