Friday 27 December 2019

Google Analytics glitch in an ad-filled internet

Ad blockers are a constant of the internet, but what does that mean for us? After comparing two different sources for one KPI, I noticed a significant difference. This piece covers the observations made about the Google Analytics glitch.

A strange disparity

Whilst collecting our monthly KPIs I came across an interesting error.

One of our KPIs, demo requests, which we track in two ways – with Google Analytics events and through a custom Zapier integration that inserts the prospects’ details into a Google Sheet and our internal Attio account.

After comparing the two datasets, I noticed a significant difference in the month’s numbers. The number of requested demos on our Google Sheet was 22% higher than the number of recorded demo requests on Google Analytics.

Unfortunately, it’s not possible to tell which demo requests weren’t recorded due to the Google Analytics glitch but we can say for certain that a significant proportion of this month’s demo requesters were somehow subverting Google Analytics.

The Culprit

What can disrupt a websites’ tracking, is used by a considerable number of people, and is applied to websites indiscriminately?

Ad blockers.

Like almost the entire population of the internet, I hate ads. Especially the in-your-face, unskippable, make-me-hate-your-website ads. So I use an ad blocker. I’ll happily disable ads on websites that serve unobtrusive and relevant ads but as a general rule of thumb, I’ll be loading a site with ads blocked. My behavior pattern is certainly common. And consequential. Since the most common permissions for ad blockers is to opt-out, rather than in, then visitors to your website will be automatically blocking ads.

It’s not until delving into this problem that I realized that it’s not just ads that your average ad blocker is blocking. Almost anything that can track or identify you is being blocked. Google Analytics included.

Our customers vs the general public

I found a variety of studies that put the percentage of internet users using an ad blocker at 20-27%. Since it’s reasonable to assume that if you use an ad blocker you’re more technologically-savvy than the average internet user, and as a power-user centric, Web 3.0, SaaS application, you’d expect our prospects to exhibit a higher percentage of ad blocker usage.

However, the similarity in our 22% dataset disparity and ~24% of global internet users blocking ads suggests that the potential customers landing on our website are no more likely to be using an ad blocker than a random selection of internet users.

It would be an interesting extension to see if the ad blocker usage changed when browsing in a professional setting.

A cause for concern

From a data analysis standpoint, it might seem like the issue of ad blockers and blocked analytics tracking would be a problem, a misrepresentation of actual events is never good surely?

In actual fact, it’s not an issue at all.

Whilst it’s true that the demo request events aren’t tracked, neither are the page view, demo form open, or demo form close events. Holistically, our dataset in Google Analytics is exactly the same, on a relative basis, as if the blocked events were also tracked.

Now you could look into differences in user actions based on their usage of an ad blocker but the likelihood of finding any meaningful correlation is close to nil. If someone installs an ad blocker, are they more likely to request a demo for a SaaS product? Potentially, but I doubt it.

The only downside to ~24% of your visitors using ad blocking software is that it reduces the absolute value of our KPIs. When comparing year-on-year growth or conversion rates this doesn’t matter, but when presenting absolute values to potential investors it would certainly be nicer to have them 22% higher.

Alex Vale leads the growth efforts at Attio, the next generation of intelligent relationship workspace.

The post Google Analytics glitch in an ad-filled internet appeared first on Search Engine Watch.



from Search Engine Watch https://ift.tt/2SsgPGN

No comments:

Post a Comment