PostHog development at ZeroPixel
PostHog is analytics for the inside of a product rather than the outside. Where GA4 is built around traffic and marketing channels, PostHog is built around users and what they do: which steps of your onboarding lose people, which feature the customers who stay actually use, and what the seventeen people who hit an error were doing beforehand.
Three things make it worth installing properly. Session replay, which turns 'the signup is confusing' from an opinion into a video you can watch. Feature flags, which let you ship a change to five percent of users and turn it off in a second if it misbehaves — without a deployment. And experiments, which tell you whether the new version actually performed better rather than whether it felt better.
The work is in the instrumentation. An event schema designed before anyone starts firing events is the difference between analytics that answers questions in a year and a list of four hundred inconsistently named events that nobody trusts. We design the schema, instrument it in the app, verify each event, and set up the funnels and dashboards your team will look at on a Monday morning.
What we set up in PostHog
Instrumentation first, then the tools that make it pay — replays, flags and experiments built on events that mean something.
Event tracking & funnels
A designed event schema instrumented across your product, with the funnels and retention views that show where users actually drop out.
Feature flags & experiments
Ship to a percentage of users, roll back instantly without a deploy, and run A/B tests that report a result rather than a hunch.
Session replay & heatmaps
Watch real sessions — including the ones that ended in an error — with sensitive fields masked so you're not recording customer data you shouldn't.
CRM & warehouse connections
Product usage joined to your CRM, billing and support data, so 'which accounts are about to churn' becomes a query rather than a feeling.
Why hire ZeroPixel for PostHog work
- The schema comes first
- Naming conventions and properties agreed before instrumentation is what separates analytics you still trust in year two from a pile of events nobody can interpret. It's an hour of design that saves a re-instrumentation later.
- Instrumented in the code, properly
- Auto-capture is a decent start and a bad foundation. We fire explicit, typed events at the exact moments that matter — server-side where it counts, so ad blockers and client failures don't quietly delete your most important numbers.
- Privacy designed in
- Session replay records real customers. We mask inputs, exclude sensitive screens, set retention deliberately and use EU hosting where data residency matters — so a useful tool doesn't become a UK GDPR problem.
PostHog — common questions
Do we need PostHog if we already have Google Analytics?
They answer different questions. GA4 tells you which channels bring visitors and what marketing is working. PostHog tells you what happens after signup — which features get used, where onboarding loses people, why a specific user got stuck. If you run a website, GA4 may be enough. If you run a product with logins, PostHog earns its place quickly, and most of our clients run both.
Is session replay legal in the UK?
Yes, with care. It's personal data processing, so it needs a lawful basis, a privacy notice that mentions it, sensitive inputs masked, and a sensible retention period. PostHog masks input fields by default and we extend that to anything else confidential. Handled properly it's routine; handled carelessly — recording card fields or health data — it isn't.
Can we keep data in the EU?
Yes. PostHog offers EU-hosted cloud, and it's self-hostable if your requirements are stricter still. For most UK businesses the EU cloud region is the right balance — no infrastructure to run, and a straightforward answer when a customer's procurement questionnaire asks where their data lives.
What are feature flags actually for?
Separating deploying code from releasing a feature. The code ships dark, you switch it on for your own team, then five percent of users, then everyone — and if it misbehaves you turn it off in seconds without an emergency deployment. They also make A/B testing straightforward, since the flag decides who sees which version and PostHog measures the difference.
How much does PostHog setup cost?
Event schema design and instrumentation for a typical product is a small-to-mid four-figure project depending on how many screens and flows need covering. It's often the highest-return week of work on a SaaS product, because it turns roadmap arguments into evidence. PostHog's own pricing is usage-based with a generous free tier we'll help you stay inside.
Beyond PostHog
A stack is a set of choices that have to work together — these are the pieces we most often pair with PostHog, and where we do it.
Building with PostHog?
Tell us what you're making — we'll reply within one working day with an honest take and a fixed quote.