Idea teardown: how we decide which cards make the cut

We get asked often: how do you decide which ideas make it into SaaS Compass?

The honest answer is that most don't. We generate a lot of candidates and reject most of them. Here's what the process actually looks like.

The first pass: can we name the buyer?

The first thing we look for is specificity. Not "this is useful for businesses" — but a precise description of one type of person who has this problem badly enough to pay.

Most candidate ideas fail here immediately. The problem description sounds plausible but the buyer could be anyone. "Companies that need better internal documentation." That's not a buyer. That's a category.

We push until we can write a sentence like: "A senior developer advocate at a Series A B2B SaaS company who is the only person managing technical docs and spends 5+ hours per week manually updating API reference pages after engineering deploys."

Once we can write that sentence, we have a buyer. Until then, we don't.

The second pass: is the gap real?

A lot of ideas describe a problem that's already solved. The founders who would benefit are already using Notion, Linear, Stripe, Zapier, or some combination of five tools duct-taped together.

We ask: what does the buyer's current solution literally fail to do? Not "it's too expensive" — that's a pricing problem, not a product gap. What specific thing happens in the workflow that no existing tool handles?

If the gap is vague, the card gets sent back for rewriting.

The third pass: where do we find them?

This is the hardest pass, and the one most idea resources skip entirely.

We don't publish a card unless we can name the actual communities where this buyer is active. Not "developer forums." The actual subreddit. The Slack workspace. The conference Discord.

Because an idea with no distribution path is not a validated idea — it's a market research report.

What almost killed our last card

Our most recent published card was almost rejected at the second pass. The first draft of the gap said something like "existing tools don't serve this use case well."

That's not a gap. That's a sentence that could be true of almost any product.

We rewrote it three times before landing on a description that named the specific workflow limitation in the category leader, the workaround the target buyer currently uses, and why that workaround breaks down at a certain scale.

The rewrite took two hours. The card is better for it.

Why we publish slowly

We publish one or two cards per week. Not ten, not twenty.

Because the value of a card is not the idea itself — it's the 40 hours of research that went into making it specific enough to act on. If we publish faster, we cut corners on specificity. And a vague card is not just useless — it's actively harmful, because it gives founders confidence they haven't earned.

The catalog is small on purpose. Every card in it passed the same three passes.