The freshness tax

There is a quiet operational cost in every SaaS company that nobody puts on a dashboard. Call it the freshness tax. It shows up every time engineering ships a change and someone, somewhere, has to rewrite a sentence about that change.

The bill goes to a few places. Support pays it when an agent reads the wrong article and gives a confident answer that's wrong. Onboarding pays it when a tour highlights a button that moved to a different menu. Sales pays it when a prospect hits a screen that doesn't match the demo. Marketing pays it when the website still describes the previous version of a feature.

The freshness tax is the gap between what your product does today and what every other system says it does.

Why it compounds

Most teams have grown numb to it. They accept that the help center will always be one step behind. They staff a docs team and treat the lag as permanent weather. They add review checklists to PRs and pretend the checklist will scale with the release cadence.

It won't. The number of artifacts you need to keep current grows with the surface of the product. The release cadence grows with the size of engineering. The two curves cross early, and after that, every ship date adds a small amount of debt that someone will eventually pay.

You can see the tax in support volume. The questions that come in spike for two days after every release, then settle into a new baseline that's slightly higher than the old one. Some of that spike is users seeing real changes. Most of it is users hitting the gap between "what changed" and "what the docs say."

The structural fix

You cannot write your way out of this. You will lose to your own roadmap.

The fix is structural. The source of truth has to move from the documents to the product itself. Every system that needs to know how the product works has to query the product, not a description of it.

That isn't an AI claim. It's an information architecture claim. If you keep two stores of truth, one will always be stale, and the staler store is the one your customers see.

Where AI does enter is in the practical question of how. A human team cannot keep up with a continuous-delivery release schedule. An agent that uses the product the way a real user would can. It clicks through the flows the same way every week and notices when the click paths change. It rewrites its own understanding without anyone filing a ticket.

This is the bet we are making with Frigade. The agent learns by using your product. The day after you ship a change, the agent has already noticed.

What this looks like in practice

Three concrete shifts happen when the source of truth becomes the product:

The support team stops triaging "is this article still right?" tickets. Confidence in answers stops being a function of how recently someone updated the help center.

The onboarding team stops shipping flow updates as a deliverable that ships with the product change. The flow updates itself, in production, in roughly the time it takes the agent to relearn the affected workflows.

The sales team stops finding out at the end of a demo that the screen they're showing doesn't exist anymore.

None of those are revolutionary on their own. Together they recover the time you've been spending paying the tax.

A specific example we keep thinking about. One of our customers told us they were planning to start retiring the help-center articles and tour content they'd built up over years, and replace those surfaces with the Frigade Assistant generating live walkthroughs that update themselves whenever the product ships. Their reasoning was straightforward: if the agent knows the workflow, the article is just a stale copy of what the agent already has. The team that used to maintain articles is already doing higher-leverage work. That's the freshness tax becoming a tailwind, in real time, in a real customer.

The asset version

There is one more thing worth saying about this, because it's the move that flips the whole picture.

For the last decade, shipping faster has been a liability for everything downstream of engineering. Every release made the docs more wrong, the tours more broken, the demos more out of date. The cost of velocity was real, and it ate into the gain.

Once the source of truth is the product, that flips. Shipping faster becomes an asset for the customer, not a liability for the docs team. The agent learns the new behavior on its own. The customer sees fewer surprises, not more.

That's the part of the freshness tax most people miss. It isn't just a cost you can avoid. It's a cost that, when you stop paying it, turns into a tailwind.

If you've been treating "keeping the docs current" as a permanent line item, the question worth asking is what your product team would build if they got it back.