Plausible review: analytics that answer the four questions I actually have

I ran Google Analytics for six years mostly because it was free and I had never questioned it. The thing that finally moved me was not privacy — it was that I could not answer “did last month’s post do anything” in under five minutes without a saved report I had built and forgotten.

Plausible answers that in one screen. That is the entire pitch, and for a marketing site it is close to correct.

Setup

One script tag. Ten minutes including deploying the change. The dashboard was live before I finished the migration checklist I had written for myself.

The GA4 import ran overnight and brought over pageviews and top referrers cleanly. Everything that depended on GA’s event schema arrived as noise. I would not plan around the import; treat it as a rough historical backdrop and start clean.

What it does well

It loads instantly and it is legible. One page, one time range, everything on it. Sources, pages, countries, devices, and a goal list. My weekly review of “what did people read and where did they come from” now takes about ninety seconds.

Goals and custom events are simple enough that I actually configured them. I have four: signup started, signup completed, pricing viewed, docs viewed. In GA4 I had meant to set these up for three years.

The public dashboard is a real feature. I share the marketing site’s stats link openly, and it has been quietly useful — people trust numbers you are willing to show.

Where it stops

This is a traffic tool, not a product analytics tool. There is no cross-session funnel, no retention curve, no way to ask what a specific cohort of users did in week three. When I wanted to know why trial conversion dropped in May, Plausible could tell me that traffic and source mix were unchanged — genuinely useful — and then it was out of answers, and I was in the database writing SQL.

That is a deliberate design decision, not a bug, and the docs say so. But it means the honest recommendation for most SaaS businesses is: Plausible for the marketing site, something else for the product. I run exactly that split, and paying for two tools is a real cost that the rating has to account for.

The other friction is pricing on pageviews. A single post going wide pushes you into the next tier for the month, which is the moment you least want a bill. It is not expensive; it is just the wrong shape for spiky content sites, and I have been bumped twice.

Support

Two emails, both answered inside a day by someone who plainly worked on the product. The second answer told me the thing I wanted did not exist and was not planned, which is more useful than a roadmap link.

The verdict

Four out of five, and the missing point is not a complaint about quality — it is scope. Plausible does one job about as well as it can be done, and the moment your question gets more specific than “where did they come from and what did they read”, you are opening a second tab.

For a content or marketing site, I would install it again this afternoon. For understanding a product, it was never the tool.