We considered subscription pricing for RevenueGuard. We modelled $79/mo with a free trial, the usual playbook. The math said it would clear several times more revenue per buyer than the $19-per-run model we shipped with. Here’s why we said no anyway.
1. The job is once a quarter, not every Tuesday
A small business runs a customer concentration check on the cadence its lender or board asks for it — usually quarterly, sometimes annually, occasionally before a covenant test. Three to four times a year is a reasonable upper bound for almost every buyer.
Charge $19 once and the buyer pays us for what they used. Charge $79 a month and they pay us for nine months they didn’t.
2. The math we ran
At $19 per run, two runs per year, we earn ~$38 per active buyer per year.
At $79/mo with the standard SaaS churn curve (free trial, 4% monthly churn after month three), the same buyer pays us roughly six months on average. That’s ~$474 per buyer.
About 12× more revenue. Per buyer.
That number should make you suspicious of every SaaS pricing page you’ve ever seen.
3. Where the extra $436 actually comes from
The buyer doesn’t suddenly run twelve concentration reports because they’re paying us monthly. They still run two. The extra revenue isn’t them buying more value — it’s them paying for the months between uses, when they forgot to cancel.
This is the dirty secret of every “transactional job priced as a subscription” SaaS. The product team’s job stops being “make the bot better” and becomes “keep the customer from cancelling.” Dashboards exist to make logging in feel like usage.
4. What we actually want
We want the buyer to come back when they need RevenueGuard, not because they’re locked in. That has implications:
- If RevenueGuard stops being worth $19, you stop running it, and we get the signal in our dashboard the same week. We can’t hide behind annual contracts.
- We can’t fund a customer success team with retention dollars, so the bot has to actually work without one.
- Per-run pricing means the bot’s quality compounds on its own merit, not on the inertia tax.
5. When we’d reconsider
If you’re running RevenueGuard against more than 50 client books a year — say, an accounting firm doing it for every retainer — that’s not transactional anymore, that’s an integration. We’d build a real API contract for it. But we’d still bill per run, just on a single Stripe invoice for the count agreed.
Email hello@botx402.io if that’s you.
If you read this far and you still want to give us money every month, please don’t. Run RevenueGuard at $19 when you actually need it. Neither of us has to pretend a once-a-quarter report is a relationship.