Plausible and Fathom are great — but they aren't attribution
Cookieless analytics is a genuinely good idea with one structural consequence: no identity means no journey. Here is what that costs and when it doesn't matter.
cookieless ⇒ no persistent identity ⇒ no journey
Plausible and Fathom are excellent privacy-first analytics: fast, simple, EU-friendly, and cookieless by design. That design choice means they cannot connect a visitor across days, or connect an anonymous session to the email it later becomes.
Not a criticism — it is the trade they consciously made, and for content sites it is the right one.
Plausible and Fathom did something admirable: they looked at Google Analytics, decided the trade-off was wrong, and built the opposite. Lightweight script, no cookies, no personal data, a single readable dashboard, and a business model that does not depend on advertising.
If you run a content site, a blog, a documentation site or a portfolio, stop reading and go install one. They are better for you than anything on this website.
The structural consequence
Cookieless means no persistent identifier. No persistent identifier means:
- A visitor today and the same visitor on Friday are two different visitors.
- An anonymous session cannot be connected to the email address it later becomes.
- A purchase eleven days after the ad click has no link back to that click.
- There is no “customer”, only sessions.
That is not a gap in the implementation. It is the direct consequence of the promise, and both products are honest about it.
The head-to-head
| Plausible / Fathom | Flowsk Signals | |
|---|---|---|
| Unit of analysis | Session (no persistent id) | Person |
| Cookies | None | First-party, server-set |
| Cross-day visitor stitching | No | Yes |
| Anonymous → email → purchase | No | Yes |
| Server-side conversion confirmation | Limited | Core |
| Client + server de-dup | No | Yes, on a business key |
| Journey for one conversion | No | Yes |
| CAC / ROAS by campaign | No | Yes |
| Consent needed in the EU | Usually not | Usually yes |
| Script weight | ~1 KB | A few KB |
| Price | From ~$9 / mo | $29 / mo |
When cookieless is enough
If your conversions happen in the same session that produced them, session-level source data answers your question. That covers a great deal of the web:
- Content and media sites measuring reach.
- Documentation and marketing sites where the conversion is “read the page.”
- Products with same-session signup and no paid acquisition worth optimising.
- Any team whose privacy position makes consent-gated measurement a non-starter.
When it is not
The moment you are paying for traffic that converts later, session analytics stops being able to answer the only question that matters. A $150 DTC purchase takes days. A B2B trial takes weeks. In both cases the session that converts is not the session that was acquired, and no cookieless tool can join them.
Run the attribution window visualizer with your own conversion delays. Everything to the right of “same day” is invisible to a session-based tool.
The honest privacy comparison
We are not going to claim parity here. Plausible’s position — no cookies, no personal data, no consent banner in most EU contexts — is genuinely simpler and genuinely stronger.
Flowsk stores an anonymous UUID and, when you send one, an email address you already hold under some lawful basis. It is first-party, it never sells or shares data, and there is no ad network on the other end of it. But it is identity-based measurement, and in most EU contexts that belongs behind consent.
That is the trade. You get the journey; you take on the consent obligation. If that trade is not acceptable for your business, Plausible is the better product and we would rather say so than pretend otherwise.
Running both
Plausible for traffic, Flowsk for revenue, is a common and sensible setup. Neither duplicates the other, both are light, and the combined bill is under $40. Plausible even makes its dashboard public if you want the traffic numbers open.
Frequently asked questions
Can Plausible do attribution?
It shows traffic sources and goal completions per session, which answers 'where did today's conversions come from' for same-session conversions. It cannot answer 'which campaign acquired this customer' when the purchase happens eleven days later, because there is no persistent identity to connect the two.
Is Flowsk GDPR-friendly?
It is first-party by design and stores an anonymous UUID plus, when you send one, an email you already hold in a lawful basis. It is not cookieless, so in most EU contexts it belongs behind consent. Plausible's position is simpler — that is exactly what you pay for it in capability.
Can I run Plausible and Flowsk together?
Yes, and it is a good combination: Plausible for fast public traffic analytics, Flowsk for attribution and revenue. Both are light and neither duplicates the other.
Which is better for a blog or content site?
Plausible or Fathom, easily. If nobody is buying anything, there is no conversion to attribute and their simplicity wins outright.
Which is better for an ecommerce store running paid ads?
Flowsk for the money question, and keep Plausible if you like it for traffic. Cookieless analytics cannot tell you which ad paid for a customer who took a week to decide.
See what a cookieless setup cannot hold
Describe how long your buyers take to convert and see how much of that distribution needs a persistent identity to capture.
Stop guessing which ad made the sale.
Flowsk Signals stitches the anonymous click to the email to the purchase — first-party, server-side, de-duplicated. One snippet, $29/mo, and every conversion comes with a receipt you can inspect.