Coverage report — example on synthetic data
Agent support-assistant · 18 Aug 2026 – 21 Aug 2026 · generated 30 Aug 2026
The tail is short
There are few enough distinct execution shapes to address all of them: there is no long tail for spend to disappear into, so there is nothing to choose between. This is the most favourable case — the 80/20 ratio does not apply here, because it assumes you have to select.
The 3 most expensive patterns — out of 5 in total — account for 90.5% of the spend, across 20.6% of the runs.
Where the spend concentrates
Cumulative share, patterns ordered from most to least expensive. The gap between the two lines is the point: a pattern can be half your runs and a rounding error on your bill.
Cost per pattern
The 5 patterns absorbing the most spend. These are not necessarily the most frequent ones: a rare, expensive pattern weighs as much as many frequent, cheap ones.
Addressable cost
The spend of the 3 patterns covering 80% of traffic.
This is not a projected saving. It is the spend a cheaper path could act on, and it becomes a saving only for those patterns that pass validation — run in shadow mode against live traffic, comparing the cheaper path with the current one and measuring how far they diverge. This report does not contain that measurement.
Cache keys you could widen
Input fields that varied while the result stayed the same. The cheapest saving here, since it needs no model.
No field could be shown inert. Every candidate varied together with request, so no pair isolates it. This is common
on agents whose input is one blob of request text, and it is a property of
the data rather than a failure of the measurement.
5 fields were shown to change the result.
Full distribution
Top twenty patterns by frequency.
| # | Pattern | Runs | Share | Cumulative | Cost | Avg cost |
|---|---|---|---|---|---|---|
| 1 | pat_9a83b666c61ed528a7f4a12d | 108 | 54.0% | 54.0% | $0.0871 | $0.00081 |
| 2 | pat_65a2c8d97ddd684b4e3d1d07 | 50 | 25.0% | 79.0% | $0.1074 | $0.00215 |
| 3 | pat_d197d790f677c8b6dcd98b83 | 24 | 12.0% | 91.0% | $0.3342 | $0.0139 |
| 4 | pat_bbf600a246b06686a4ac3567 | 13 | 6.5% | 97.5% | $1.04 | $0.0800 |
| 5 | pat_3b1f9616a2d53b6c222623da | 4 | 2.0% | 99.5% | $0.4821 | $0.1205 |
Models and tools
Where the spend goes, by model invoked.
| Model | Calls | Cost |
|---|---|---|
| claude-haiku-4-5 | 159 | $0.1952 |
| claude-sonnet-5 | 24 | $0.3342 |
| claude-opus-5 | 17 | $1.52 |
Most invoked tools: crm_lookup (122), ocr (50), erp_write (50), kb_search (37), ticket_update (13), sql_query (4), chart (4)
What this report does not tell you
- Costs are estimated from a price table where the provider does not report them: they indicate the order of magnitude, not an invoice.
- A frequent pattern is not compressible by virtue of being frequent. Its runs must produce equivalent outcomes, and this report does not measure that.
- The observed period may not be representative: seasonality, spikes and prompt changes all shift the distribution.
- Grouping is by semantic cluster. A finer grouping may split patterns that appear merged here.