See what we're building
Pineapple on pizza

No data available

Feature ownership
Members

Roadmap

What we're building

  • AI Wizard Installer

    Install PostHog with a single command, rolling out to all users soon!

    Project updates

    No updates yet. Engineers are currently hard at work, so check back soon!

  • Vercel Integration

    We're building an integration with Vercel so that you can sync your Feature Flags and Experiments to your NextJS deployments.

    Project updates

    No updates yet. Engineers are currently hard at work, so check back soon!

Goals

Q2 2026 objectives

North star: 20% of sign ups should happen outside posthog.com/signup

These are the primary goals the team is prioritizing this quarter.

Make agentic sign ups first-citizen of PostHog (Matt Brooker Fernando Gomes)
  • Rationale: Agents are increasingly the ones picking and setting up tools. If our sign-up assumes a human clicking through a dashboard, we miss the fastest-growing top of funnel.
  • Things we could do: Agentic onboarding by default — headless onboarding, Wizard, Stripe projects, AI-first flows. Wizard should be able to create orgs + projects on its own with no UI clicks. Expand the pattern beyond Stripe (done) to Vercel, Lovable, Supabase, and Replit.
  • We'll know we're successful when: An agent can bring a PostHog org and project online through any of our partner surfaces without a human in the loop.
Allow anyone to integrate/provision an org + project in PostHog (Rafael Audibert Matt Brooker)
  • Rationale: Every partner, every embedded platform, and every agentic flow needs a programmatic way to spin up a PostHog org or project. One-off integrations don't scale — a generic path does.
  • Things we could do: Lean on the soon-to-come Partnerships Wrangler. Market it as either PostHog Marketplace or PostHog as a platform. Review what Agentic Provisioning needs to become generic (just an API). Write docs and settle on how to do it. Stand up the right application process — probably Support Hero-driven.
  • We'll know we're successful when: A partner or agent can provision an org + project through a single public path, with no custom work from us.
Activation rate across all onboarding experiences should be 60%+ (Matt Brooker Fernando Gomes Rafael Audibert)
  • Rationale: A single numerical bar pulls every onboarding experience towards the same outcome. Different channels — warm, lukewarm, cold, frozen — need different treatment, and a unified target forces us to treat them as one system.
  • Things we could do: Settle the activation metric alongside PMs — we're currently on "soft activation" per-product; global activation is still TBD. Implement channel-specific onboarding: website (warm), Wizard (lukewarm), agentic provisioning (cold), Lovable/Replit auto-adding Analytics (frozen). Explicitly de-scope PostHog Code activation for now — Signals + PostHog Code team will handle that.
  • We'll know we're successful when: Measured activation rate across every onboarding surface is 60% or higher.
Shape the external interface for Signals (Matt Brooker Fernando Gomes Rafael Audibert)
  • Rationale: Signals is still early and its external surface isn't settled. Getting the shape right now — before partners depend on it — avoids painful contract breaks later. Partner vibecoding app builders are the most credible early validators.
  • Things we could do: Write a spec for the external interface and validate with partner vibecoding app builders. We believe these partners want Signals, but Signals may not be ready to expose yet — first check feasibility. Once that's done, stop and reflect: how do we make this the next big thing after MCP?
  • We'll know we're successful when: We have a validated spec and a clear read on whether Signals can be the next big thing after MCP.

Stretch goals

Q1 2026 recap

  • Onboarded Fernando Gomes to the team
  • Shipped MCP Apps
  • Shipped Stripe projects
  • Shipped Vercel integration
  • A bunch of MCP improvements alongside the PostHog AI team
  • Shipped VSCode extension
  • Shipped HogSpy
  • Reverse proxy revival (mostly Daniel, but we made it more relevant in-app)
  • MCP got featured as connector in a bunch of different marketplaces
  • ...and much more, look at the changelog

Next quarters

Non-exhaustive list of things we want to work on eventually:

  • One-click migration away from competitors
  • Demo environment for customers to try out PostHog with existing data (and relatable to their own business domain)
  • PostHog-billed AI SDK
    • Don't need to setup an AI account in each provider, just use PostHog's AI SDK.
  • Improved ingestion URL (dynamic, harder to block/guess)
  • New products playground

Handbook

What the Growth team does

We operate under a triple mandate - albeit one easier than the FED's: increase top of funnel acquisition, ensure successful activation, and increase cross-sell between products.

  • Activation
    • Global onboarding flow
    • Develop tooling for individual product onboarding and activation
    • Help product teams leverage the tooling to create best-in-class onboarding & activation flows for their products
  • Product led growth
    • Develop new products and strategies to increase top of funnel
    • Integrate with external partners to drive growth
  • Product cross-sell
    • Track and measure cross-sell between products
    • Identify the right time to cross-sell different products to different users
    • Work with marketing to experiment on cross-sell initiatives outside the app
    • Build bridges to other products that appear when customers might benefit from them

How to work with the Growth team

In addition to covering specific areas of the product, members of the

Growth mini crest
Growth Team
Growth mini crest
Growth Team can temporarily join other small teams to execute jointly on growth-related tasks. This can be done by requesting help during Sprint planning, or at the beginning of the Quarter during the OKR definition phase.

Feature ownership

You can find out more about the features we own

Questions about this page? or post a community question.