Data Differentials: The Parallel Universes of AdWords & GA | DMA

Filter By

Show All
X

Connect to

X

Data Differentials: The Parallel Universes of AdWords & GA

T-558969d888dae-star-trek_558969d888cc2-2.jpg

On our mission we have many awesome tools at our disposal. Arguably the two most important are AdWords and Google Analytics. Sometimes scanning both systems something seems amiss. Could it be we are looking at our data reflected in a different way? As if reflected in a mirror, perhaps you might even say a mirror universe? Maybe your data has a funky new goatee.

Mirror Universe Spock With A Wicked Goatee

So what are the main reasons for the difference between AdWords and Analytics data?

The main reason for the discrepancy we see between AdWords & Analytics is the attribution model used differs slightly on each platform. In other words both systems are actually tracking different things. Its life Jim but not as we know it.

Its Life Jim But Not As We Know It

AdWords attributes conversions to the last AdWords click, which means that if a user clicks an AdWords link at any point in the conversion path this is attributed as a conversion in AdWords. Following this system, if the user visits our site via a PPC link and then leaves, but later returns and converts (inside the 30 day conversion window) via an organic listing, AdWords would count this as a conversion, but it would not appear as a conversion under AdWords in GA.

For some accounts this can happen much more frequently than others. For example the setup of Google shopping encourages users to shop around more, comparing prices, models and delivery charges. This is especially true in verticals where a large number of competitors chase every single sale. Exactly the verticals where Shopping is most commonly used.

In many cases users will visit a site and also visit competitors before returning to purchase on the site offering the competitive terms or best service. This can result in significantly more revenue being shown for shopping campaigns in AdWords versus Google Analytics where the conversion might not be attributed to shopping.

You may be able to see evidence of this in the multi-channel funnels section of Google Analytics.

Let’s take a look at some examples.

Multi Channel Funnel Showing Significant Overlap

Showing paid search in blue, notice there is significant overlap with both organic search in green and direct in the brown on the path towards conversion.

The multi-channel funnel top conversion path report gives you the best indication of the paths users are taking to conversion.

Multi channel funnels popular paths

In this example the majority of converting users enter the site via paid search and then return at a later point directly and convert. Only the paths that end with a paid search click would be attributed to AdWords in the Google Analytics multi-channel funnels report.

While AdWords takes the credit for every path to conversion where AdWords features, GA really gives us a fuller picture of the importance of paid search, as it's possible to see all the interactions between AdWords and other channels.

There is also a difference in how Google Analytics attributes conversions between different reports. In the multi-channel funnels report the last entry point will always be shown at the end of the sequence, whether that is Direct, Organic or Paid Search. In other GA reports, conversions are attributed slightly differently, as the last non-direct entry will win the conversion. In the above example, numbers 1, 3, 4, 5, 8 & 9 would be attributed to Paid Search, 6 & 10 to Organic Search and 2 & 7 to Direct (as direct was the only touch point in those two cases).

Note that within the attribution modelling tool reports you can choose the attribution model that is used. If you use the Last Click model, the very last click will get the credit but if you use the Last AdWords Click model, the data should be within Google's recommended tolerance of 10% compared to AdWords.

As you can see in the above example, where and how you view the data can lead to significant reporting differences.

Other Common Reasons For Data Discrepancy

There are other reasons why there might be differences in conversion data between AdWords and Google Analytics, and often this is a result of the differences between the two products in the way that clicks and sessions are tracked. Here are some of the most common reasons for these discrepancies:

Browser preferences. Some users may have a cloaking device. OK, ok, more accurately they may have JavaScript disabled on their browser. This would enable AdWords to log the click but not Analytics as the JS Analytics cookie will not fire.

Server latency. In some cases a user may click an AdWords link and immediately hit back or close the page before the Analytics JavaScript has the opportunity to fire. Again AdWords will record a click but Google Analytics will not record a session.

Clicks are not sessions. If a user clicks your ad and visits the site this is both a click and a session. However, if the user leaves the site, only to return via an ad within 30 minutes this would be another click for AdWords but would be part of the same session for Analytics.

Time (and Space). AdWords attributes conversions to the date of the click, whereas analytics would attribute that same conversion to the time the goal action was completed. In some cases, for example an expensive purchase requiring significant consideration, this may be a large difference.

Invalid click activity. Some goal completions shown in analytics may be filtered out in AdWords by Google's Invalid Click technology which won’t record suspicious or invalid clicks.

Cookie Expiry Date. AdWords cookies expire after 90 days and generally have a 30 day conversion window, whereas an analytics cookie expires after up to 6 months. If a user completed a conversion action after the AdWords expiry it wouldn’t be shown in AdWords but would still be visible in Analytics.

Goal Import Delay. This is a doozy. Analytics goal completions take about 48 hours to show as a conversion when imported into AdWords. This will always give you a discrepancy when comparing recent AdWords and analytics data. Due to this, I would always recommend the use of the AdWords Conversion tag wherever possible over tracking using GA goals and importing back to AdWords.

I hope with the help of Mr Spock we have all learned a little bit more about the final digital frontier, and pulled back the curtain between AdWords & Analytics. Live long and prosper. Peace and long life.

Leonard Nemoy

Leonard Nemoy 1931-2015

To view the blog as written by James Olver on Periscopix's website, please click here.

Hear more from the DMA

Please login to comment.

Comments