Why does cookie-free analytics measure more traffic, not less?
Because the consent banner is the leak. Google Analytics sets cookies and collects data that identifies individuals, so under the GDPR and ePrivacy rules it needs explicit consent — and every visitor who clicks reject vanishes from the numbers entirely. Depending on the audience, that is routinely a third of traffic and sometimes far more.
Plausible, Simple Analytics, Pirsch and Counter.dev set no cookies and collect no personal data, so no consent is required. That is a technical fact rather than a legal interpretation: there is nothing to consent to. The banner disappears, and with it the sampling bias that comes from measuring only the people who agreed to be measured.
The bias is not random either, which is what makes it dangerous. Privacy-conscious visitors decline more often, and they are disproportionately the technical, high-intent audience many businesses most want to understand. A dashboard built on consented traffic systematically underrepresents exactly the segment worth knowing about.
Pirsch is explicit about the mechanism it uses instead: a daily hash of IP address, User-Agent and a salt identifies a unique visitor within one day and is regenerated every day, so nobody can be followed across days or identified at all.
What do you give up by leaving Google Analytics?
Depth of user journey analysis, and it is worth being honest about how much of that you were actually using. These tools give pageviews, visitors, bounce rate, referrers, countries, devices, goals and custom events. What they do not give is funnel analysis, cohort analysis, user segmentation or multi-touch attribution modelling.
Simple Analytics states the limitation plainly: no funnels, no segmentation, no cohorts, and limited e-commerce tracking. Pirsch adds that daily hash regeneration means individual journeys cannot be followed across days at all — a direct consequence of the privacy design rather than a missing feature.
For most websites that is a trade worth making, because the funnel report was opened twice a year and the numbers underneath it were already missing a third of the audience. For a business whose revenue genuinely depends on multi-touch attribution across a long sales cycle, it is not.
None of these seven offers heatmaps, session recordings or A/B testing either. Those are separate products, and Pirsch says so rather than pretending otherwise.
How much does the analytics script cost your page speed?
The Google Analytics tag is roughly 45KB of JavaScript that has to be fetched, parsed and executed before it reports anything. Plausible's script is under 1KB — around a fiftieth of the size — and the others in this category are comparably small.
That matters beyond the abstract. Core Web Vitals feed into search ranking, and a heavy third-party script competes for the main thread exactly when a page is trying to become interactive. Removing 44KB of blocking JavaScript from every pageview is a measurable improvement, not a rounding error.
There is a second effect that is easy to miss: a smaller, single-purpose script is far less likely to be caught by ad blockers and tracker-blocking browsers, which quietly delete a further slice of Google Analytics data before the consent banner even appears.
So the performance argument and the measurement argument point the same way. The lighter script is also the one that actually runs.
Cloud or self-hosted — which should you actually pick?
Self-hosting removes the processor entirely. Plausible is AGPL-licensed with Docker deployment scripts, and the self-hosted version is functionally identical to the cloud one — not a crippled community edition. Pirsch is AGPL v3 and self-hostable too, and Counter.dev is MIT-licensed. For an organisation with data sovereignty requirements written into a contract, that is the only configuration that satisfies them.
The cost is the usual one: a server, updates, backups and someone responsible when the dashboard stops collecting. At €9 a month for Plausible cloud or €4 for Pirsch, that person's time is worth more than the subscription for any small team.
Simple Analytics has no self-hosted option at all. Only the tracking scripts are open source; the application itself is proprietary. If code transparency or self-hosting is a requirement, that rules it out regardless of its other merits.
The middle position most teams land on: use the European cloud, and treat the open licence as the exit route if the vendor is ever acquired or changes direction.
What should you check before switching?
That the goals you actually report on can be recreated. All of them support custom events and goals, and Plausible adds custom properties, but the definitions do not migrate — someone has to rebuild them and confirm the numbers line up.
That you can live without the historical comparison. Nothing here imports Google Analytics history, so the year-over-year comparison resets. Running both in parallel for a quarter is the usual answer, and it also quantifies exactly how much traffic the consent banner was costing you.
That the reporting reaches the people who need it. Simple Analytics offers public dashboards, which several companies use to show traffic to advertisers, investors or customers without granting account access — a genuinely useful feature that has no equivalent in Google Analytics.
And that the pricing model matches your traffic shape. Pirsch includes unlimited websites and unlimited team members on every plan with no feature gating, which suits an agency far better than per-site pricing does.
