Skomi and PostHog

Nothing is stored on the visitor's device, by default. PostHog uses cookies unless cookieless mode is switched on, and when it is, it identifies visitors with a daily server-side hash, which is the same technique Skomi uses. The difference is which way round the default points, and a default is what most installations keep.

There is no instrumentation project first. PostHog rewards a team that models its events carefully, and it depends on one. Skomi's page views, sources, goals and funnels work from one script with nothing named in advance.

It is a smaller thing to run. PostHog's breadth is real and it is also surface area: more products, more settings, more to configure and to keep configured.

It is open source and Skomi is not. MIT licensed, and you can host it yourself, with the caveat PostHog states plainly, that self-hosted deployments are officially unsupported and get no paid support plan. If inspectable code matters, this is the only product here that answers it fully.

Feature flags and experiments, which Skomi has none of. Shipping behind a flag and measuring the result in the same tool is a real advantage, and it is not a gap Skomi intends to close.

Surveys, error tracking, a data warehouse. Each is a product Skomi does not have. If you would otherwise buy three tools, PostHog replacing all of them is a stronger argument than anything else on this page.

A free tier most teams never leave. Generous enough that price is simply not the deciding factor for a lot of companies.

Choose PostHog if you want open source, feature flags and experiments, deep product analytics, or one platform covering far more than measurement.

Choose Skomi if you want site and app measurement that needs no consent banner by default, no event modelling before the first useful number, and one bill covering analytics, behaviour and feedback.