PostHog tells you everything. Most of it is noise.

Pick a B2B SaaS company at random and reconstructing its roadmap takes a week of careful reading. Pick PostHog and they hand it to you — a public page, organised by team, with the things they decided against still sitting there in the open.

Which should make it the easiest competitive intelligence job in the category. It isn't. Radical transparency doesn't remove the work, it relocates it. When a company publishes tens of thousands of words about itself, the scarce resource stops being access and becomes judgment. So here is a teardown run in reverse: not "what can we find," but "of everything they've already given us, what is actually load-bearing?"

The handbook is a recruiting document in a strategy costume

The company handbook is the surface everyone points at first. It's long. It covers how teams are structured, how compensation is reasoned about, how decisions get made, what the company believes about remote work and hiring bar and meeting culture. It reads like an internal document that leaked on purpose.

Treat it as a competitive artifact and you'll come away with very little. Handbooks describe intent, and intent is the cheapest thing a company produces. What the handbook does tell you — reliably — is the shape of the org. PostHog organises around small autonomous teams, each owning a product area. That's a structural fact, and structural facts are durable. The count and naming of those teams is a better predictor of where engineering effort lands next quarter than any roadmap prose, because a team that exists has salaries attached to it and a team that doesn't, doesn't.

So read the handbook once, extract the org shape, and stop. Then check back only when the team list changes. A new small team appearing is a real event. A paragraph about culture being rewritten is not.

The roadmap is real. The discard pile is realer.

The public roadmap board is the surface that earns its reputation. It's genuinely the thing itself, not a marketing rendering of the thing — issues, owners, states, and public discussion attached.

But the highest-signal column on any public roadmap is the one nobody screenshots: the requests that were considered and closed. When PostHog declines a feature in public, with a reason, you're reading an actual constraint — a market they've decided isn't theirs, an architectural cost they won't pay, a customer segment they're happy to lose. That's strategy expressed as subtraction, and it's far harder to fake than a "coming soon" card. We've made the broader version of this argument before: a competitor's rejection pile is the most honest part of their roadmap.

The trap with a board this active is volume. Dozens of items move every week. If you're tracking all of them you're tracking none of them, and you'll burn a researcher's attention on card churn. Watch two things: items that move from "considered" to "in progress" without passing through public discussion (someone senior wants this), and items that sit in progress for two quarters (something is harder than they expected).

The pricing calculator is the product portfolio

PostHog's pricing page is usage-based with a calculator, and the calculator is the most underrated surface they publish. Not because of the numbers — because of the list.

Every product with its own meter in that calculator is a product the company has committed to operating as a separate line: separate infrastructure cost, separate free tier, separate expansion story. Products that share a meter haven't earned that status yet. When a feature graduates from "part of the platform" to "its own row with its own price," that's a portfolio decision that happened months earlier in a room you weren't in, surfacing publicly for the first time.

The generous free tier does the same work in the other direction. A free allowance that's large enough to run a real startup on isn't pricing — it's a distribution strategy that says the company expects to win on land-and-expand rather than on procurement. That tells you exactly how they'll show up in your deals: bottom-up, through an engineer who already installed it, not through a VP who ran an evaluation. Pricing pages encode org structure and strategy more reliably than about pages do — PostHog's just encodes it more legibly than most.

Your competitors aren't PostHog. Invert the exercise.

Here's the part that transfers. Almost nobody you compete with publishes like this. But every surface above has a quieter equivalent at a normal company:

• Org shape — the careers page. Team names in job titles do what a handbook's team list does.
• Discard pile — the community forum or GitHub issues, where "won't fix" and polite closures live.
• Portfolio commitments — the pricing page's SKU list and the nav bar's product menu.
• Distribution strategy — where the free tier ends and the sales-assisted motion begins.

The reason PostHog is worth a teardown isn't that they're a likely competitor for you. It's that they're a calibration target. Read the company that publishes everything, note which of those surfaces actually predicted what shipped, and then go look for the muffled version of exactly those four surfaces on the companies that publish nothing.

This is roughly the job Seeto does on the ongoing side: it monitors those public surfaces continuously and surfaces the differences as discrete change events — a new pricing row, a changed team list, a nav item that appeared. It won't read the handbook for you or tell you what a change means. Nothing automated should claim to. But it does mean the four surfaces above stay watched between the quarters when you actually have time to think about them.

The most transparent company in your market isn't giving you an advantage. It's giving you a training set.