With Google's sunset date of Universal Analytics (UA4) coming up next week, I figured it would be a good time to detail one of the often recommended alternatives — Plausible Analytics — and how I'm doing 6 months later as a paying customer.
I run the tech stack at YIP, which includes our production website. While not directly impacted by these changes, we were planning a domain migration for the start of 2023 regardless and I saw the timing as an opportunity to move away from our implementation of Google Analytics (GA). We were running into several pain points with GA, notably the script size and lack of real-time data. Therefore, after soliciting feedback we began looking for something that met the following criteria:
- Lightweight. Our web host, Webflow, leaves us with little control over the HTML/CSS/JS served to users and reducing the size of scripts was one of the few things we could manage.
- Privacy First. We only really needed view counts, general location, and basic attribution (?ref=xyz) tracking - not a 30 day cookie like Google Analytics.
- User Friendly. In addition to the tech team as stakeholders for a replacement, we had several non-technical users who just wanted the stats on a specific page.
- Cost, the ability to export data, and being open source were all other criteria identified but were not dealbreakers.
After some research in December 2022, we had several routes to explore:
- Drop-in replacement. Simple Analytics, Fathom, Matomo, and Plausible all market themselves as alternatives or replacements to GA/UA4.
- Development around existing platforms. We use Dub for link shortening, and they use Tinybird for the underlying tracking.
- Bespoke installation of Enterprise software. Two products, Mixpanel and DataDogHQ, came across our radar but were ruled out due to the amount of development work needed for an install.
The platform I found that fit our needs the most was actually Vercel Analytics, but this was incompatible with our web host and there were no plans to migrate at the time. Our timeframe was closing in to make a decision, as this was an afterthought for our primary goal of changing our domain for the new year to .org. Therefore, we explored the "drop in" replacements and reached the following conclusions:
Simple Analytics fit our privacy needs and opened us up to the concept of cookie-free tracking, but an emphasis on data security meant it was difficult to share that data with other stakeholders in our organization.
Matomo was heavily geared towards the European market and out of our price range.
Our last two contenders offered a live demo (Fathom, Plausible) of their reporting dashboards and this was enough for us to do a trial install on a subset of the website. Plausible ended up being a better fit, had faster customer service for a question we asked both providers, and a lower entry price. These factors, along with being open-source (and able to understand where development was headed) were enough to sway us into using Plausible as a replacement. I would say "Christmas came early!", but it was December 28th at this point.
Installation and Aftermath
Permanent installation of Plausible went smoothly, and we began collecting data on December 29th ahead of the domain migration on the 30th (also went smoothly!) The main feature we were missing was being able to compare across periods beyond top-level views/visitors, although this has since been added. Several other things happened after our shutoff of GA:
- No more automated weekly reports! We had a slack channel using Narrative BI to alert us about traffic spikes and provide weekly reports, but Plausible only offers reports in the form of weekly and monthly top-level snapshots you can have emailed to addresses of your choosing. As a workaround, we setup the Slack channel's email address as a recipient of these reports; these reports are still a little rough around the edges.
- No more Data Studio! Our Data Studio reports became frozen in time at the end of 2022, and no effort was made to update them. Plausible doesn't offer first-party connector to Data Studio but there are 3P options out there for a price. We can export data as CSV or use the API, but this was not an internal priority.
- Search Console got mad that it didn't have accurate data anymore, but still provided impression insights so it was fine with us. Plausible offers a linking option to Search Console, which pulls in the keywords visitors searched that attributed clicks, but it's not 100% accurate.
Overall, the team was very happy with the rollout and it was one of the better parts of our domain migration. Our favorite and most used features include:
- Location information for visitors - supported at the country level worldwide and city/metro area in many countries too.
- Tracking outbound clicks for normal links and file downloads.
- A larger emphasis on inbound sources, as opposed to GA where this is buried inside reports. Combined with Plausible's simple dashboard, this makes diving into why a particular source is or isn't driving traffic much easier.
- True "real time" tracking updated every 30 seconds, comparable to what GA offers.
- Support for ?utm=xyz tracking along with the cleaner ?ref=xyz tracking for social media posts.
- The ability to add a CSS class to an interaction (eg. a button) and track the # of clicks that button gets for old-school conversion tracking on forms.
The one area we are still struggling with is granular access; we can only share the whole dashboard with users (or guests), with no options to restrict by path. A workaround would be to setup different tracking tags on each page and create a "rollup" dashboard, but this would create problems every time we release a new page. This was not an area GA could really compete in natively, so I would consider the migration a resounding success.
This post isn't sponsored or anything, but personally I really like the open-source yet financially secure approach Plausible has taken with their product. For the least friction, you can use their JS snippet at a fair price; or, self-host using your own server with a moderate amount of effort. Either way, their Github serves as a roadmap for upcoming features and an opportunity to see what other product users are doing for inspiration. YIP is a satisfied customer, and with Plausible disclosing $1 million in ARR we weren't the only ones looking at the UA4 migration as an opportunity to de-Google a bit :)
Until next time,