You've rewritten the homepage three times this year.

Each version felt better than the last. The copy is crisper. The hero is cleaner. The value prop is more specific. And yet — something still isn't landing. The pitch still changes depending on who's in the room.

This isn't a writing problem. It's a canon problem.

What canon means here

In storytelling, canon is the agreed body of truth. The events that actually happened. The rules of the world. When writers on a show disagree about whether a character would do something, they go back to canon.

Commercial systems need canon too. The agreed set of decisions that everything else is built from.

When these decisions are made and documented, they function as canon. Every surface — homepage, deck, outreach, demo script — gets built from the same source.

When they're not made, every surface gets rebuilt from scratch. By whoever is writing it. With whatever they remember from the last conversation.

Why positioning slides aren't canon

Most founders have a positioning slide in their deck. It has a category, a one-liner, maybe a competitive matrix.

That slide is a summary, not a canon. It captures a conclusion without the evidence and the reasoning that produced it. When someone new joins the team, they see the conclusion but not the decisions underneath. When the market shifts, there's no document to update — there are just twelve places where the conclusion got embedded in slightly different words.

This is why positioning drifts. Not because people forgot. Because the decisions were never written down as decisions — they were written down as outputs.

What locked decisions look like

A locked decision isn't "we serve mid-market B2B SaaS companies."

It's: "We serve mid-market B2B SaaS companies with 50-200 seats, in their second product cycle, where the founding team is still running sales and feels the drag. We know this because our first five customers all had this pattern, and the two we lost were both earlier stage — no sales drag yet, no urgency. We're committing to this segment and saying no to enterprise for now because the sales cycle would break our current model."

That's a decision. It has evidence. It has a rationale. It has a choice that was made. When someone new reads it, they understand not just what you decided but why — which means they can apply it to novel situations.

The compounding problem

Here's what happens without canon:

A new sales hire joins. They study the deck. They start selling from the slide. But the slide doesn't answer the objections they're hearing on calls. So they improvise. Their improvisation is different from what the founder says. Now you have two versions of the story in the market.

Then you write a content piece. It uses slightly different framing than the sales pitch. Now there are three versions.

Each piece of drift makes the next piece of drift easier. The internal bar for "good enough" keeps moving because nobody's measuring against a fixed point.

The canon is that fixed point.

How to build it

The Sprint process builds canon in a specific sequence, because sequence matters.

First: the customer decisions. Who, precisely. What they're using instead. What job they're trying to get done. These decisions come first because everything else depends on them.

Second: the differentiation decisions. What you do that nothing else does. How to frame it against the alternative. What the mechanism is. These decisions can only be made once the customer decisions are locked — differentiation is always relative.

Third: the proof decisions. What claim you can commit to. What evidence supports it. What risk exists and how to address it. These decisions depend on knowing who you're making claims to.

Fourth: the action decisions. What you ask them to do first. Why that step and not another. How you price it. These decisions depend on all the above.

When you do this in the wrong order — which most positioning processes do — you get conclusions that don't hold. The one-liner doesn't survive the first objection because the underlying customer decision was never actually made.

What changes when canon exists

The homepage rewrite stops being a creative project and becomes a translation task. You have decisions. You're translating them into language that a skeptical buyer can read in ninety seconds.

The onboarding script stops depending on institutional knowledge. A new hire can read the canon and understand not just what to say, but why.

The pitch stops changing per presenter. Everyone is working from the same source.

None of this happens automatically. Someone has to run the process. But once the canon exists, maintenance is much cheaper than creation. You're updating decisions at the margin, not rebuilding from scratch every time something shifts.

The sprint locks the initial canon. Everything after that is refinement.