Guidelines & Best Practice

Why clicks reported in an adserver will never match click source reporting in Google Analytics

Mon, 10 Oct 2016


By Toni Knowles, GM NZ, VeNA

7th October  2016

Being immersed in the digital advertising industry, which has evolved to a level of maturity but also increasing levels of complexity, it is easy to lose sight of the fact that there are many new advertisers who are just now expanding their digital advertising and marketing beyond the self service advertising platforms of Google and Facebook.

Outside of these ‘walled gardens’ is the ‘open web’ which presents a massive opportunity for advertisers – publishers and mobile apps with quality content and helpful utility for consumers to engage with, and at the same time presenting new and valuable audiences for advertisers.

These new advertisers may be entering into a relationship with an advertising agency for the first time, or expanding their own marketing efforts in house to engage with major digital publishers and new digital advertising distribution platforms.

There are also many marketing and advertising professionals getting involved, who have not yet had the chance to get up to speed on some of the questions that we all experienced early on in our digital marketing journeys.

One age-old question that popped up again for us this week is, Why don’t the clicks in the adserver reporting match our clicks showing in Google Analytics?

There are many reasons why these two data sets won’t match – here are a few of them:

  1. First and foremost, they are two completely different events. A click reported on an ad, and a user reported as having landed on a page of a site, with the web page fully loading and firing the analytics are not the same event.
  2. They are also measuring different metrics – an adserver tracks all clicks, vs google analytics tracks sessions. For example, if a user clicks on your ad twice within thirty minutes without closing his or her browser, this is registered in the adserver as two clicks, but in Google analytics as one session, even if the user left your site and then returned shortly after. Another example, if a user clicks on your ad once, clicks the back button, and then clicks your ad again, the adserver registers two clicks while Google Analytics registers one session.
  3. There are different counting methodologies to consider – time zones, date ranges of reporting may not match the configuration of the ad server data vs the google analytics reports.
  4. Time outs – an adserver may have recorded the click, however the user navigated away from the landing page, before the landing page fully loads and records the visit, because the page takes too long to load. Additionally across the adtech ecosystem there can be multiple tags firing all within milliseconds, that get counted, but again the user has navigated away prior to Google analytics recording the visit (eg the ad creative and the google analytics URL may be hosted in an agency side server, which is passed through the programmatic advertising technology layers, then to the publisher adserver before rendering to the end user). Time outs can sometimes be more pronounced in the mobile landscape. As more audiences migrate to mobile, it’s super important to have a mobile optimised website and fast page loads.
  5. Accidental clicks on the ad by the user – the click will be reported in the adserver, but the user intentionally navigates away before the landing page loads.
  6. Incorrect google analytics set up or reports – if you’re looking at referring domains, not all domains can be identified. Using UTM tracking strings will help identify the source of the visit, but ensure all pages where you’re directing traffic have google analytics tags implemented correctly.
  7. Incorrect implementation of the google UTM tracking URL in the adserver – if the URL is not entered with the query string intact, or with missing info / white space, the UTM code may be stripped out by the user’s browser and therefore not count in the google analytics reports
  8. User browser settings – some users will choose not to be tracked. The clicks will be reported in the adserver, but not as a known source in Google Analytics

Other considerations:

  • The brand value of video advertising in particular is widely proved to be effective (top of the funnel, consideration stage, awareness) – ie it’s not all about the click
  • All adservers have filters in place to filter out any invalid or suspicious clicks and impressions. The counting methodology of adservers is reliable.
  • A publisher would never be able to investigate a discrepancy between their reported clicks and google analytics reports – it’s just not possible as the metrics are two different things. They will be more focused on ensuring their data will closely match the agency’s third party adserving data reporting – even then, these numbers won’t match exactly – the accepted variance across the industry is that the data should be within 10-20%. Impressions will usually be within these variances (the impression counting methodology of adservers is largely similar). Clicks can sometimes be more disparate as adservers may use different filtering and counting of clicks.
  • For advertisers new to all of this, it is worth investing in some expertise help in getting Google analytics configured correctly to get the best use from the tool, a greater understanding of what the data showing in the reporting represents, and which metrics to focus on for understanding the campaign’s performance in relation to their specific goals.