Naia

So here's the real question we're trying to answer with Naia — can it grow how much clients value us and stay with us, faster than it grows our AI costs? That's the whole game.

And notice what we're not saying: we're not trying to reprice everything overnight. We just want to learn.

The reason we're being careful is there are four traps we already know are out there. First, the “paying twice” feeling — clients already license this data, so asking them to pay again just to ask a question can rub the wrong way. Second, cost that's hard to predict — our costs move with usage, but client budgets are annual and fixed. Third, margin leakage — a handful of heavy users could quietly rack up real cost. And fourth, cannibalization — position this badly and Naia makes our existing products feel less essential instead of more. Keep those four in mind, because everything after this is about avoiding them.

So how might we actually sell this — and how do we figure it out without just guessing? On the left are the commercial models on the table.

Included credits — a set allowance each month or year — is our first pick to test, because it feels fair and it's predictable. Tiered access, where premium tiers get higher limits, is promising too and fits nicely with how we already package and upsell. Credit packs are useful, but more as a top-up for heavy users than a main model. And pure pay-per-prompt we'd rather avoid as a default — it just makes every question feel like it costs money.

Now, the plan to learn all this is deliberately simple. We run a beta with a couple of hand-picked clients who'll give us honest feedback. Naia's included at no cost during the beta, we set a soft usage allowance and watch it internally, and we learn what people actually value and what it really costs per account — before we put a price on anything. The goal isn't perfect pricing on day one; it's understanding what's fair to both sides.

And that brings us to the questions worth sitting with — as a group. None of these need an answer today, so please don't feel put on the spot.

These are just the things the beta is designed to help us work through together — across product, data, and how we eventually take Naia to market. Where does it create the most value: which workflows, and which people, does it genuinely make better? What feels fair to charge for: how much should feel “included”, and which model sits comfortably next to what clients already pay us? And where might clients hesitate: what framing or pricing could cause friction, and how do we get ahead of it?

No decisions here — just worth keeping these in the back of your mind as the beta plays out, because that's exactly where the real answers are going to come from.