Coverage and credits
Coverage is a wallet of answers, in two pools, allocated each period. Here is what spends it, what happens when it runs out, and how to make it go further.
Your plan allocates coverage each billing period: a wallet of answers, split into two pools. Everything the platform does on your behalf spends from it, and the spend is visible rather than implied.
What you'll learn
- The two pools and what draws on each
- What a check costs, and why costs differ
- What happens at zero, and how to stretch the allowance
The two pools
| Pool | Pays for |
|---|---|
| Monitoring | The ordinary work: checking your prompts on each covered surface, including AI Overviews reads and product checks |
| Generative | The expensive work: grounded checks, content briefs and forecast simulations |
Splitting them protects the routine. One enthusiastic afternoon of brief generation cannot silently consume the monitoring that your reporting depends on.
What a check costs
Metered actions carry a weight in answers, derived from what the action actually costs to run. A plain citation check on one engine is the cheap end. A grounded check, a content brief or a forecast simulation is the expensive end.
The published rate card is at pricing/rate-card. Read it before you plan a large prompt set or a batch of briefs.
The metered actions are: citation checks, grounded checks, retrieval checks, AI Overviews reads, product checks, content briefs and forecast simulations.
Where it goes
The dominant factor is almost always prompts multiplied by surfaces multiplied by cadence. A prompt is checked on every surface your plan covers, every cycle, whether or not it is a good prompt.
Which is why the honest first response to running low is rarely an upgrade:
- Retire dead prompts. Anything that has never produced a citation or a competitor sighting.
- Check the intent mix. Informational prompts are the easiest to write and the least valuable to hold.
- Be deliberate about products. A product check meters per product, per engine, per window, so a whole catalogue is not a thing to track exhaustively.
- Then upgrade, if the remaining set genuinely needs more.
Rollover, top-ups and trials
- Rollover. Unspent subscription coverage can carry into the next period up to a cap that depends on your lane. It is a cushion, not a savings account.
- Top-ups. Additional coverage can be added when you need it.
- Trials. A trial wallet is deliberately smaller than the full lane allowance; the full grant lands when the trial converts.
At zero
Monitoring pauses rather than quietly degrading, and the account says so. Two distinct events exist, and the difference matters:
- Coverage exhausted. You spent what you bought.
- Coverage not allocated. A lane has no allocation at all, which is a billing fault on our side, not you spending.
Both are available as webhooks, so your own systems can know before a client does.
For agencies
The wallet is pooled across the whole agency, not rationed per client. That is usually an advantage, and it means one client with an enormous prompt set can consume the pool for everybody. Size the allowance before switching a large client on. See agency billing model.
Try this in Outercite
Open usage and look at the split between the two pools. If your generative pool is untouched, you are not using briefs or simulations, and if it is empty early every period, that is the thing to plan around.
Related
Plans and limits
What each lane allocates.
Add prompts
Where most coverage goes.
Webhooks
Being told before a client notices.
Last updated on