JF Martin writes Numeric Citizen. On 27 January 2021 he published Here is why I Switched from Google Analytics to Plausible. This is his own post, not a vendor page.
He had used Google Analytics since 2010, first on Blogger, later on his blogs. After a 30-day Plausible trial he dumped GA.
If you are looking at the same route, start with the Google Analytics → Plausible playbook. JF’s post is the privacy-and-simpler-reports field report. The playbook is the import, dual-run, and verification path he barely describes.
Why he left
He likes stats. He used them at work in datacenters. At home he watches iStats Menu on the Mac. He wanted the same for the blogs: what is going on, without a lecture.
Google Analytics had been the default since that 2010 Blogger site. He said it was time to walk the talk on privacy. GA was not privacy-friendly. That is his reason. He does not publish a GDPR letter or a cookie-banner fight.
WordPress already showed visitor numbers. GA on top of that felt redundant on the main blog. The microblog had no built-in stats at all. He wanted one tool that covered both.
He also started 2021 with what he called a digital reset. Leaving GA was part of that, not a traffic emergency. He does not publish a GA bounce-rate complaint or a bill. GA was free for him.
Why Plausible
He ran the 30-day trial, then stayed.
Plausible is open source. Setup on two blogs took less than five minutes. Support answered fast. The metrics were thinner than Google Analytics. He said they were more than good enough.
He named the gaps. Plausible was web-only in January 2021. There was no official API, so no native iOS app. Graphs were “ok”, not as flexible as he wanted. He still called it good enough.
The price was $4 a month, billed annually. That is his 2021 number. I am not updating it to today’s Plausible list price.
He also flipped the dashboards public. Plausible still documents that toggle. Stats start private. You turn on Visibility and get a shareable URL. He said it was not about bragging. He wanted the numbers in the open, the same way he already wrote about Google algorithm swings.
A year later, in Friday Notes #68, he still pointed people at those public Plausible stats. He listed Plausible next to Ghost, GitHub, Micro.blog, and WordPress as the open tools he pays for on purpose.
How he moved
He does not describe an export, an importer, or a dual-run spreadsheet. The post is a 30-day trial, then a cut.
Adding the script is a snippet in the page head. He already ran WordPress on at least one blog. Plausible now has an official WordPress plugin with a token connect, a proxy toggle, and a stats screen inside wp-admin. He never names the plugin, so I am not saying he used it. Five minutes on two sites also fits a pasted snippet.
He does not mention importing GA history. Plausible can import GA4 aggregates now. That importer brings day-level totals, not visitor journeys. Unique visitors over a month are a sum of daily uniques. Ecommerce revenue does not come across. Goals have to be recreated in site settings before they show up. None of that is in his post. The import is extra work if you still want last year’s shape.
He also does not say how he removed the GA tag. The playbook wants a dated parallel window, then one owner. He only says the trial convinced him, and GA went away.
What he gained
He paid for a tool that matched the privacy stance he already wrote about. The two blogs shared one kind of dashboard. The microblog finally had numbers.
He cancelled the GA-only extras. Numerico had iOS 14 widgets on top of Google Analytics. Databox pulled GA into other dashboards. Both went. He deleted the widgets and cancelled Databox.
The public dashboards stayed a habit. In 2022 he still invited readers to look. I am not quoting today’s visitor counts from later dashboards, because those are not in the 2021 post.
He does not publish a before-and-after traffic chart or a page-speed score, so I am not inventing either.
What broke or got skipped
He lost the breadth of GA graphs and the Databox layer that sat on them. He said he did not need that breadth.
He also lost a native iOS stats app, because Plausible had no official API yet. That part aged. Plausible now ships a Stats API. In March 2023 he wrote that he wanted an iPhone widget through Shortcuts and that API. I did not find a later post where he shipped it.
He does not describe a failed install, a missing goal, or a rollback. Plausible’s own check is the thing I would run after the snippet lands: real visits show up as soon as they are counted, and a banner tells you if the script is missing.
Historical GA reports stayed in Google unless he exported them. He never says he did.
What I would copy
Run the trial on every property you actually care about. He put two blogs on Plausible for 30 days before he cancelled anything.
Write down the GA-only tools first. Widgets and Databox died with the property. If a weekly report still reads from GA, that report dies too.
Decide whether you need the old totals. If year-over-year charts matter, use Plausible’s GA import and label those days as imported. If you are resetting on purpose, skip the import and start clean, which is what his post reads like.
If you already run WordPress, use the plugin so the snippet, the proxy, and admin exclusion live in one place. Then confirm a real visit in the dashboard before you delete the GA tag.
Run the Google Analytics → Plausible playbook. Use JF’s post for the small, privacy-first, paid-tool-replaces-a-free-one path.