Does Google AdWords Have a Synonym Problem?

What’s the difference between ‘web design’ and ‘website design’? Or ‘SEM agency’ and ‘search engine marketing agency’? To humans, there is little difference; however, we’re seeing some interesting AdWords Quality Score variations between these similar keywords, even when taking other factors into account like CTR, CPC, landing page content, and ad copy.

The differences in Quality Score seem to be directly tied to the keywords in Ad Groups where we’re buying terms that are fairly synonymous.

Are Synonyms Google AdWords’ Kryptonite?

If Google’s Quality Score is actually impacted by terms like ‘web design’ versus ‘website design’ in a meaningful way, the implications are twofold.

First, advertisers are going to be forced to create more landing pages. Landing pages definitely have their place when you have groups of keywords that are fundamentally different than other groups. However, creating two separate landing pages for ‘web design’ and ‘website design’ is basically useless. Creating additional content because Google struggles to delineate these terms is a waste of time and resources and will further clutter the web.

Secondly, it means users will get less relevant results. We all know Google has one goal – relevance. That’s why Quality Score revolutionized paid search. While Bing and Yahoo! couldn’t avoid the race to the bottom with the traditional keyword auction model, Google realized that paid search needed relevance incentives. Even though they have evolved the model, we’ve noticed there could be an issue in assigning appropriate Quality Scores to synonyms.

The Hypothesis

We believe that Google assigns higher quality scores to keywords based heavily on the exact variation of the keyword being included in ad and landing page copy. Even if two keywords are synonymous, such as ‘web design’ and ‘website design’, Quality Scores will be higher for ‘web design’ than ‘website design’ if the ad and landing page copy only uses the term ‘web design’ – even if CTR and conversion rates are higher for ‘website design.’

The Test

Over the next few months we’re planning to run several tests to see if this is actually an issue or just an anomaly. Although we can’t be sure that the tests will bear conclusive data, we are interested in seeing how things shake out.

We’ve already run tests where we’ve dynamically inserted synonyms into ad and landing page copy, but we think Google may be too smart (or dumb) for this.

So, we’re going to build out Ad Groups where we use the exact same ad and landing page copy, for two very similar keywords. For example, we will test "web design" and "website design" in their own ad group with the same ad copy/landing page copy - copy that is focused on ‘web design’, not ‘website design’. After these have run for a significantly relevant period, we’ll move 'website design' out into its own ad group with updated ad copy and a new landing page focused on ‘website design’. We’ll then compare the differences in QS, making sure to factor in CPCs, CTR, and other QS factors.

Once tests are finalized, we’ll share additional details about the experiment on our blog. As the tests are running, we’ll share the results and findings as well.

What Do You Think?

While we’re conducting our tests over the next few months, we’d love to hear if anyone else has any insights or actual data regarding this issue. Do you have a synonym problem? Feel free to add your comments or join the discussion on Facebook or Twitter.

Written by Brian Easter on September 27, 2011


Add A Comment

My head is a bit spinning. Not clear what you are trying to do, what Google is trying to do... I would simply concentrate on scope of the landing pages, write some appropriate content and not bother about Google all together. But this is just me...
Or is it just about split testing? Then grammar and layout could be more relevant...

Beaster 1 6apepxm
Written by
Brian Easter