Ingest
The file arrives in CSV, Excel, JSON, or PDF. Gate one: does it contain names, seat counts, spend, and usage metrics? Pass means the file enters. Fail means it returns with a missing-fields list. No partial ingest.
a fraction of what a top-tier firm would charge.
Fixed at $50,000.00
AI Spend VerdictSample Deliverables
The pipeline
The file arrives in CSV, Excel, JSON, or PDF. Gate one: does it contain names, seat counts, spend, and usage metrics? Pass means the file enters. Fail means it returns with a missing-fields list. No partial ingest.
Every tool is scored against a primary source ranked in order — vendor admin export first, published industry study second, nothing else counts. Gate two: does a rank-one or rank-two source exist? Pass means the cell scores. Fail means not scored, with the exact missing item named. Two sources conflict, higher rank wins, conflict logged.
Three outcomes only. Keep, shrink, not scored. No fourth option. The active-to-seat ratio at or above ninety percent is keep. Below thirty percent is shrink. Between thirty and ninety is shrink-with-a-question — flagged for CFO review but still emitted as shrink. A tool with fewer than ten active users and no cost data doesn't get a verdict. It gets flagged insufficient data.
Insufficient data
It is not a failure of the audit. It is a finding. The CFO learns which tools their own teams couldn't produce records for, and that absence is the conversation.
The deliverable
The one-page memo: total spend, the keep, shrink, and not scored split, and the top three CFO questions.
The full row-level table: every tool, its ratio, its source rank, and its verdict.
The renewal-readiness summary: which flagged tools are up for renewal inside twelve months.
The fee
Two thousand sourced entries across fifty categories. Primary-source verification on every scored cell. A top-tier firm engagement runs three to five times this, covering a narrower slice.
Start the audit