author-avatar

Julien is the analytics manager and has a passion for problem-solving. When not compiling data or discovering trends, he likes to count the trees that go by while on his mountain bike, or skis.

Other posts by:

Google Analytics or Google AdWords: Which Data Should I Trust?

Why You Can’t Compare your Google Analytics and Google AdWords Data

So you’ve decided to launch a PPC campaign. Great choice! You’re likely using AdWords and leveraging all of the latest trends, and because AdWords is part of the Google family you can link it up to Google Analytics (GA) as well! But wait. Why is the data different between platforms?

Google Analytics vs Google AdWords data

Here are a few things to keep in mind before comparing the figures from each platform (spoiler alert, they won’t always match).

Google Analytics and Google AdWords Pull and Treat Data Differently

Assuming that all data is equal across each platform is the most common mistake. You have your AdWords clicks on one side and your GA sessions on the other, and they are not matching up. GA can sometimes show 25% less sessions than AdWords clicks, which is actually quite normal because each platform is measuring it differently:

  • AdWords records a ‘click’ when a user clicks on the ad leading to the site.
  • Analytics records a ‘session’ when the GA code on the website actually loads (which is why we always advise implementing the code in the  <head> of your website).

If a user closes your website as soon as he clicks on the ad (because of a miss-click on a mobile device for example), AdWords will record 1 click while GA has no session to associate with it. Also, if your website has its GA code at the bottom of the page, you will miss out on some users that bounce. GA will simply have no record of them.

The way that each system measures performance is fundamentally different, so how does that affect the data integration between them?

Google Analytics Pulls Some Data From AdWords and Vice Versa

When your accounts are linked, you will see cost, clicks, and impressions in GA. These numbers should always match their AdWords counterparts because they are pulled directly from the AdWords API! Even when your cost is adjusted after fraudulent clicks, the data in GA will reflect those changes. This is all fine and dandy, but it gets a little tricky when we talk about conversions. AdWords pulls conversion data from GA the same way that GA pulls cost, clicks, and impressions from AdWords, but this time the roles are reversed.

First off, please make sure all the conversions you want to track are enabled in AdWords. To see which ones are and aren’t, head over to “Tools > Conversions” to configure. Also make sure that you are comparing the same conversion types in both reporting instances as this might be why your data doesn’t match up. Even after this is done, you will still likely get some discrepancies between the two different conversion data sets.

Why you ask? GA keeps all data. Everything. Whereas AdWords deletes what it considers to be “fraudulent activity”. If a conversion is associated with an AdWords click that is suspected to be fraudulent, both the click and the conversion are removed from the data. This happens most often for easy to achieve goals (destination page views, max scroll, etc.), which is why we encourage the tracking of more down funnel goals like ‘send an email’, ‘click to call’, ‘form submission’ and so on.

AdWords Has Unique Conversion Types

Another reason you may be seeing discrepancies is the fact that some conversation types are only trackable/available in AdWords. For example, Google Forwarding Call completions and Google Sponsored Promotion conversions are not available GA. These only show up in AdWords.

The Attribution Model and Lookback Window Are Different

By default, the attribution model used by both systems is “Last Non-Direct Click”; however, there are several attribution models available on both platforms, so it’s important to make sure you’re comparing the same model in both.

The lookback window is different by default in each tool, and this is where it gets tricky. By default, GA is set to a lookback window of 6 months, while AdWords is set to 30 days. You can adjust AdWords’ lookback window to 90 days, but that’s obviously not the same as 6 months. The affects the attribution because the lookback window is used by Google to match up multiple visits from the same user. If one exists in that time, it will be included in the attribution model. If not, there is no difference. Here’s an example to help explain this:

  1. User 1 comes to the site in January from an AdWords click.
  2. He then comes back directly to the website in April because there is a sale on his favorite item.
  3. He buys it!

In GA, the revenue will be attributed to ‘CPC’. In AdWords, the revenue will be attributed to ‘Direct’. Fun!

Now What?

Hopefully after reading this post, you will look at your data in a different light. Here are the things to remember when you do go dig in your numbers:

  1. Pull website conversions from GA.
  2. Pull AdWords specific conversions (GSP, Calls) from AdWords.
  3. Pull clicks, impressions, and cost data from GA and reate a custom report for your campaigns, this will make your life a lot easier.
  4. Pull attribution data from GA.F

All things considered, the best practice is pull as much data as possible from GA because it can give you valuable insight into user behavior and conversions. AdWords is still the main source for all bid adjustments and keywords, but if you want to use it exclusively you’re most likely running a CTR or CPC model. To get to the next level, you need GA!

Now that you know what you can and can’t pull from AdWords and GA, have fun with your data! And please comment with any of your concerns and questions for us! Or send us an email! We’re more than happy to help out! a-team@vovia.com