Head to head
Umami vs Plausible Analytics
Free and yours, or paid and polished. Scored on ten weighted measures, with every price read on the same day.
Umami
7.5/10
No. 4 of 30
Plausible Analytics
8.0/10
No. 2 of 30
Both are open source, both are cookieless, both can be self-hosted, and both give you a single clean dashboard with no consent banner. The difference is where each one puts its effort. Umami is small, MIT licensed, and easy to run on your own Postgres. Plausible is a six-year-old business with twenty-one thousand paying subscribers, and its product shows that investment.
The verdict
It depends on your budget and your appetite for operations.
Take Umami if the site is personal, a side project, or documentation. Its hosted Hobby tier gives you a hundred thousand events a month for nothing with no badge and no expiry, and self-hosting is genuinely an afternoon: a Docker deploy against Postgres or MySQL, far lighter than the ClickHouse cluster Plausible's community edition expects.
Take Plausible if the site earns money. The depth gap is real: saved and shared segments, automatic scroll-depth tracking, AI traffic reporting, Google Search Console integration, funnels and user journeys on its Business tier. Umami's funnels and breakdowns are competent rather than rich, and its bot filtering is less developed than the tools that make a feature of it, which means raw counts deserve more scepticism.
A note on the upper tiers: Umami's session replay and heatmaps only appear at two hundred dollars a month, which is not a realistic route to behaviour tooling for most readers.
Side by side
| Umami | Plausible Analytics | |
|---|---|---|
| Overall score | 7.5 / 10 | 8.0 / 10 |
| Rank on this list | No. 4 of 30 | No. 2 of 30 |
| From | Free self-hosted and on cloud Hobby | £9 per month |
| What that buys | Cloud Hobby is free at 100,000 events a month; Pro is $20 for 1M events | 10,000 monthly pageviews, on a Starter tier |
| Free tier | Yes, 100,000 events a month on hosted cloud, plus free self-hosting | No, but the community edition is free to self-host |
| Cookies | None | None, so no consent banner |
| Script size | About 2 KB | About 1 KB |
| Hosting | Your own servers, or hosted cloud | EU, on European-owned infrastructure |
| Trading since | 2020 | 2019 |
| Free option | Self-host free; 100k events/mo hosted free | Self-host free; no free hosted tier |
| Self-hosting stack | Node with Postgres or MySQL | ClickHouse cluster |
| Entry paid tier | $20/mo for 1M events, 20 sites | £9/mo for 10k pageviews, 1 site |
| Segments | Basic | Saved and shared segments |
| Search Console integration | Not offered | Yes |
| Heatmaps and replay | Business tier only, $200/mo | Not offered |
| Bot filtering | Present, less aggressive | Built in, including datacentre traffic |
The ten measures, scored
| Measure | Umami | Plausible Analytics |
|---|---|---|
| Insight quality | 7 | 9 |
| Privacy architecture | 9 | 9 |
| Value for money | 9 | 8 |
| Setup and time to data | 6 | 9 |
| Script weight | 9 | 9 |
| Analytical depth | 6 | 7 |
| Behaviour tooling | 4 | 2 |
| Data ownership | 10 | 8 |
| Accuracy | 6 | 8 |
| Vendor conduct | 7 | 7 |
| Weighted overall | 7.5 | 8.0 |
Higher is better. The overall figure is the weighted mean, relative to the 30 tools on this list. How this works.
Which one should you pick
Choose Umami if
The site is personal or a side project, or you want an easy self-host on a database you already run.
Choose Plausible Analytics if
The site earns money and you want deeper segmentation, Search Console integration and a vendor with a track record.
Common questions
Which is easier to self-host?
Umami, by a distance. It is a Node application against Postgres or MySQL. Plausible's community edition expects a ClickHouse cluster, which is a meaningfully larger operational commitment.
Is Umami's free hosted tier genuinely usable?
For a small site, yes: 100,000 events a month, one site, six months of retention, no badge required. The retention window is the real limit, since six months cannot support a year-on-year question.
Which gives more accurate numbers?
Plausible, slightly, because its bot and datacentre filtering is more developed. Both avoid the consent-banner losses that affect cookie-based tools, which is the larger effect by far.
Can I get heatmaps from either?
Only from Umami, and only on its $200 Business tier. Neither is a realistic route to heatmaps for a small site; a tool that bundles them from a low tier is the cheaper answer.
Both entries above are desk reviews written from published evidence, and every judgement is my opinion rather than a statement of fact. Read the full reviews for Umami and Plausible Analytics, or tell me where I have got something wrong.