Announcement came in April 2022 that, in July 2023, Google will retire Universal Analytics and the Reporting API that Post Pay Counter PRO has so far used to retrieve visits data. This is not our choice: Google is simply taking away the way you have used to track visits so far, and the mean we have used to fetch them into PPC so far. Significant backlash has not lead them to reconsider the choice. With 6 months to go, we obviously have to ponder about this as well.
I assumed the new GA4 would be an improvement over the existing service, and that adapting an old implementation shouldn’t require too much work other than renaming some function calls. This was grossly mistaken. The whole story is a shitshow. I wrote a post about all the ways in which, as a developer and user, GA4 is , I recommend you have a read even just to understand the efforts I have put into not disrupting the service PPC currently provides. To everybody who asked, I have always been optimistic that PPC would eventually support GA4, and the migration would cause no pain to our customers. I am not so sure this is the case anymore. Skimming over a lot of details that you can find in the blog post, the main points are:
- GA4 does not even have a stable API yet. I brought the issue up once and twice to Google. The only reply I got over months is blame on some other Google team. As if I cared. They are totally deaf to users’ needs.
- GA4 does not allow you to import data from your old Universal Analytics views, and UA views will be inaccessible in 2024. This means that you won’t be able to run long-time Analytics data imports into PPC, as it is now possible.
- GA4 properties cannot be linked to Adsense, which means our current Adsense integration will break with GA4 views. At the same time, it simply does not seem possible to use Adsense to retrieve the revenue data we now take from Analytics. not. Request for support has, time and again, fallen in a black hole.
Overall, the user and developer experience of GA4 is appalling and frustrating, for a myriad of other small (sometimes technical) reasons. I’ve already put it many efforts in replicating the current PPC flow in GA4, to no avail. And if, as it looks from the points above, supporting GA4 is basically like supporting a new visits system (since no historical data is available), then I’ve come to question whether that is even worth it. Shall we not support new platforms that are modern, better documented, with better support, and that are a tenth of the pain for me to maintain? Maybe a platform whose docs are decent enough that I don’t have to reverse-engineer their API in order to use their library? A platform that allows me to surf through their pages without hitting a 404 every third link I click? Confronted with all these issues, Google was just silent.
These days I am less and less optimistic that PPC is ever going to support GA4. What is sure is that I am not going to put in any further work until they have a stable GA4 API and library. And even when that happens, I will ponder on this. But one thing is clear: whatever I decide, your current visits/revenue payment flow will break from July 2023, if you use Google Analytics. And there is nothing I can really do to prevent it.
(To be honest, it is also our responsibility to oppose this. If Google has the feeling that they can do whatever they want, and that the developers will comply, we can quite easily show that this is not how other human beings ought to be treated by… ditching their product!)
As the situation stands, I suggest you take action to possibly replace Google Analytics as your only measure of visits, and that you do it now. Post Pay Counter does already support several WordPress visits counting plugins. The plan is to add support for more, so that you have a wide replacement choice. It comes a bit as a surprise (but it shouldn’t) that all serious alternatives are paid. But hey, even if we’d like to think that open source developers love to work for free (which they don’t), cloud infrastructure alone has a cost, and analytics platforms receive as many hits as the sites they monitor, so they need loads of resources. Nobody can afford to offer such a service for free, other than giants like Google that do it a loss. So be prepared to pay, or to get a scam. This is a trend in tech.
I am keen on recommending an open source project that you can self-host, if you so wish, or buy a cloud instance with a monthly fee. It was my hope to recommend Matomo (formerly Piwik), which I knew from before, but it turns out it is a Frankenstein piece of code, with a poor API and documentation. It’s a pity, because they recently launched a WordPress plugin that allows to install Matomo within a WP site.
After careful consideration and research, I decided to recommend Plausible. Some points of interest:
- it is open source, and it provides a self-hosted version that is very easy to set up, with good docs
- it offers a paid cloud solution. I wouldn’t define it as cheap, since it’s as expensive as a cheap hosting, but it’s on par with current competitors. It offers a 30-days free trial without credit card, and a 2 months discount on yearly plans
- it has a WP plugin for integrating it within your site
- it is a simple service with a minimal impact on page load time, check out their live demo. Of course it doesn’t match the complexity of Google Analytics, but 80% of people only care about basic metrics anyway
I have given this a lot of thought, and I am convinced this is the right path. If you have a small site, and $10 extra a month are of concern, then you can use one of the simple free visits counting methods that PPC already supports. If your site is instead of large scale, either you can ask your developers to set up a self-hosted solution, or you can likely afford $10 extra a month. At the same time, I am very open to hearing from you all. What is your stance on this? Is there some other analytics service that you think we should support, or that should be our first recommendation instead of Plausible?