Which tools does P8 connect to?

P8 currently connects with Slack, Fathom, Google Analytics, and PostHog, with more integrations on the way. Not sure what you can connect? We’re currently offering a free personalised setup with the founders. We’ll help you work out which of your existing data sources can work with P8 and get everything connected for you so you can start seeing value without the technical hassle.

Does P8 replace the PM?

No. P8 surfaces and ranks what customers are telling you, you make the call. It's more signal, not a stand-in for judgment.

What if my data isn't clean?

P8 reads calls, tickets, and changelogs as they are. No migration, no cleanup sprint.

How is this different from tools that just summarize feedback?

Those tools stop at collection and synthesis. P8 starts where they stop, with a ranked decision.

The latest on turning signal into decisions

Notes on turning signal into decisions — pieces from the P8 team on product, growth, and how the weekly brief comes together.

A Competitor Used Your Exact Word. Here's the Check That Matters — A competitor's stealth launch used our exact positioning word. Here's the one check that separates a real threat from a name that just sounds like one.
What AI Should Actually Find in Your Customer Calls — Most "AI tool to analyze customer calls and tickets" searches lead to a transcription or tagging layer, not a decision. Here's what actually closes the gap between what customers say and what you build next.
Evidence-Based Product Management Sounds Obvious. Here's What It Costs. — Every team says they're evidence-based. Almost none of them can show you the evidence behind their last three shipped features. Here's what the practice actually requires, and why most teams quietly drop it under deadline pressure.
We Tried Feedback Tools to Make Better Decisions. Here's What We Got Wrong — A feedback repository is not a product decision system because it only performs the first step of the job: capturing and organising what customers said. It has no mechanism for weighing conflicting signals against each other, recording what you chose and rejected and why, or checking whether the bet paid off. A decision system adds the four things a repository was never built to do: reconciliation, decision capture, outcome tracking, and decision memory.
What a 5,800-Commit Release Tells You About the AI Tools Race — Hermes Agent just shipped named bots, group chats, and terminal pets in one release. None of it makes anyone's product better. Here's how to tell when a competitor's feature velocity is a signal worth reacting to and when it's just louder.
How Do You Decide Which Customer Feedback to Act On? — Every piece of feedback feels urgent when the customer is on the call with you. Most of it still shouldn't change your roadmap. Here's the filter we run each request through before it earns a place on the list.
How to Decide What to Build Next After Getting It Wrong for Years — Most teams treat the backlog as the decision. They rank tickets, sort by effort versus impact, and ship. Then they discover they built the wrong thing. The backlog is a memory aid, not a priority queue. The actual decision happens earlier, in four places: Who is this for? What outcome are we chasing? Where does it hurt in their journey? What evidence do we actually have? Get clear on those four things, fill in a one-page decision brief, and suddenly ranking becomes trivial. The hard part—deciding what matters is already done. This is how teams stop shipping the loudest request and start shipping what moves the needle.
How to Prioritize Feature Requests From Customers Who All Disagree — Feature requests don't arrive ranked, and counting votes just rewards whoever asks loudest. Here's how to rank them by customer weight, tied outcome, and source count instead.
How to Prioritize Product Features When You Don't Have Enough Data — Most prioritization frameworks assume a sample size you don't have. Here's how to make a defensible call on what to build next when you're working with twenty customers, not twenty thousand, and frequency counting breaks.
How to Prioritize Your Roadmap Without a Product Manager on the Team — Most prioritization frameworks assume a PM to run them. Here's what to do instead when the founder, an engineer, and a spreadsheet are the entire product function.
How to Run Product Discovery When You're the Only One Doing It — Product discovery frameworks assume a researcher, a PM, and a backlog of interview time. Here's how to run continuous discovery alone, with the customer conversations you're already having and none of the ceremony.
How to Spot Churn Signals in Customer Conversations Before They Cancel — Churn risk shows up in what customers say weeks before it shows up in a usage dashboard. Here's what to listen for in calls and tickets, and why counting complaints misses most of it.
A Framework for Turning Weekly Evidence Into Decisions — Most Monday reviews turn into a recap of everything that happened last week, which feels thorough but decides nothing. This piece breaks down a capped framework, three things to build and one thing to stop, that forces real trade-offs and ties every move to evidence, an owner, and a way to check later if the call was right. Includes a copyable template.
The Problem With Engagement Data: A Single User Can Change the Story. — Our CTA-click chart looked healthy for weeks. Then we split it by visitor and found one recurring, likely-internal ID behind most of it. Here's why aggregate engagement numbers lie by default and what to check before you trust one.
The Product Decision Making Framework That Actually Fits a Small Team — RICE, ICE, and Kano were built for teams with thousands of data points and a PM to run the spreadsheet. At sub-50-customer scale, the math breaks before the decision does. Here's what to score instead, and why.
Looking for a Productboard Alternative? Check Your Team Size First — Productboard is built for 100+ person PM orgs with dedicated framework owners. Here's how to tell if you actually need an alternative, or just a different kind of tool entirely.
How to Know If Your Content Is Working: Publishing Is Only the Start — Hitting publish feels like progress, but a page with zero indexed queries and a 100% bounce rate isn't evidence of anything except that it exists. Here's why treating publication as the finish line breaks the same evidence loop every other product decision depends on, with a worked example from our own numbers.
Your Impressions Are Climbing But Your Ranking Is Getting Worse — A page can show more impressions every week and be doing worse, not better. Rising visibility with a falling average position usually means you're getting matched to broader, less relevant searches, not winning the ones you wrote for. Here's how to tell the difference and what to check first.
What Not to Build: Use ERIC Before Your Next Sprint — Product improvement is not a synonym for adding features. ERIC — Eliminate, Reduce, Increase, Create gives elimination and reduction equal standing with building. Here's how to apply it to a swollen backlog, with a decision canvas.
What Should You Build Next? Here’s How to Actually Know. — "What should we build next" feels like a backlog question but it's actually three different questions wearing one trenchcoat: what's urgent, what's strategic, and what's provable. Answer them separately or the loudest one wins by default.
What to Build Next When Your Customers Want Different Things — Three paying customers, three different feature requests, one roadmap slot. This piece separates priority disagreements from direction disagreements and gives you a reach/recurrence/reversibility grid for weighing conflicting requests without just picking a favorite customer.
The System We Built to Stop Data From Telling Us What We Wanted to Hear — Three signals disagreed calls, tickets, and analytics all told a different story about onboarding. This piece rejects the instinct to rank evidence sources by authority and instead hands operators a seven-lens worksheet (identity, journey, behaviour, severity, commercial relevance, corroboration, reversibility) for weighing conflicting signals case by case. Includes a reusable reconciliation table and grounds the 'data eliminates, it doesn't decide' principle in both Nielsen Norman Group's triangulation research and Haja Deen's Build the Right Thing.