Quick help: I’ve been hacked

The statistics provided by Google Analytics are becoming increasingly less realistic

GIGA4
We are Roberto and Francisco, a small team dedicated to web development and specialized in the technical side of the WordPress and WooCommerce platforms. We provide consulting and training, we develop plugins and themes, we optimize, we secure, we respond to security incidents and we maintain sites. Founded in 2012, we are the technical support behind more than a hundred web projects in Spain and Germany, and we are also an active part of the WordPress community.

This is due to GDPR and the tracker blockers included in many privacy-focused browsers.

There are many reasons why you might be seeing low traffic in Google Analytics; for example, a textual AI suggests these reasons, which I find very accurate.

However, other reasons exist.

Tracker Blocking by Browsers and VPNs

Browsers have been adding features to block trackers (except Google Chrome for obvious reasons, unless you install an extension). Many VPNs also include this functionality.

In Spain, with data from February 2023, Google Chrome accounts for 72%, Safari 15%, Firefox 3.1%, and others share the remaining 10%.

On the other hand, a less rigorous study indicates VPN usage at 4%.

Safari has a system to block trackers, Firefox also does, and VPNs usually include it.

If these systems (or other extensions) are active on your visitors’ devices, that visit will never reach Google Analytics.

This already means a significant number of users will not appear in your statistics, and furthermore, it’s a biased profile, as they will largely be iOS, macOS, or GNU-Linux users and/or those concerned about privacy.

But there is another major filter: the cookie notice.

Safari privacy report for iPhone.

The Cookie Notice

That cookie notice that few configure correctly, because it should block the loading of scripts that affect user privacy until the user has accepted. If this notice functions as required by law, it is most likely the point where you lose the most visit registrations in Google Analytics, simply because many users reject this notice.

Fortunately, Google Analytics has a mode that records the visit without affecting user privacy, Consent Mode, which allows it to load before the user has accepted. However, its functionality is not excellent, and from the tests we have conducted, it seems to improve registration but we believe not sufficiently. Furthermore, this implementation has a certain technical complexity that prevents many from using it.

User Privacy-Respecting Analytics Systems

For some time now, we have been installing privacy-respecting analytics systems on the new websites we develop. These work alongside Google Analytics when the client also wishes to maintain that system.

In our experience, we observe that these sites receive between 30% and 85% fewer visit registrations in Google Analytics compared to other analytics systems.

This is due to everything we have discussed previously, although it must also be acknowledged that other analytics systems may not be as effective at blocking bots as Google Analytics systems, and therefore may register more visits than they should.

We have conducted tests with a site that, on the same day it generated about 160 leads, registered:

  • With Koko Analytics: 2100 visitors, 3100 views
  • With Burst Statistics: 1440 visitors, 2680 views
  • With Google Analytics: 480 users, 820 views

Which figure is more realistic? Honestly, no matter how much I think about it, I have no idea! The only thing that seems off to me is Google Analytics; having 160 leads with 480 users would be a 33% conversion rate! That doesn’t seem realistic.

The rest seem plausible. If I had to intuitively explain the difference between Burst and Koko, I would say it’s possible that Koko’s data is inflated by bot traffic not filtered by the tool. But I have no evidence to support this claim.

Perhaps the time has come when we should not read statistics as absolute values but as relative values. And not measure ourselves so much by visits, but by other reference values (leads, contacts, sales, other actions).

Burst Statistics: It has a very refined interface that provides interesting information. However, it stores every action of every session and does so with duplicated information, which over time can disproportionately increase the database size.

Koko Analytics: Although it may be slightly less accurate in recording visits and provides less data, it is very, very, very efficient at storing information (it does not save sessions or actions, only totals), so it is a system that can function for years.

Google Analytics is a fantastic, excellent system with many wonderful options and integrations. However, you should already know that if you are not paying for that system with your money, you are paying for it with your data. I believe the future is increasingly leading us to a world where more and more people are concerned about privacy, and this traffic registration gap in Google Analytics makes it increasingly detached from reality.

Leave a Reply

Your email address will not be published. Required fields are marked *