Three scenarios

Each case below ran through PreDealCheck with synthetic customer data. The engine recomputed claims from source tables and returned one of four honest states. All data is synthetic.

Scenario 1: Inflated traction

A marketplace seller claims 45 active customers and USD 156,000 annualized recurring revenue. The engine found both claims contradicted by source data.

Claim	Stated	Found	Gap	Status
Active customers	45	35	−10 (22%)	CONTRADICTED
Annual recurring revenue	USD 156,000	USD 134,460	−USD 21,540 (14%)	CONTRADICTED

The customer table held 40 rows. Rows with status "paused" were excluded as unrecognized. From the 35 customers with explicit live status, the engine summed monthly recurring revenue and annualized it. The arithmetic is transparent: row numbers, source hash, and the rule applied are attached to each finding for verification.

Scenario 2: Clean books

A SaaS founder reports 18 active customers and EUR 40,080 annual recurring revenue. The engine confirmed both claims against source data with no discrepancies.

Claim	Stated	Found	Gap	Status
Active customers	18	18	0	SUPPORTED
Annual recurring revenue	EUR 40,080	EUR 40,080	0	SUPPORTED

Honest numbers clear as readily as inflated ones. The engine distinguished customers with explicit live status, summed fixed recurring monthly rates, and annualized the result. Byte-identical match to the claim.

Scenario 3: Valid count, unsupported metric

A marketplace operator states 28 customers and claims 3 churned in Q2 2026. The customer count is valid; the churn claim cannot be verified from source data.

Claim	Stated	Found	Status
Active customers	28	28	DEFINITION_MISMATCH
Churn in Q2 2026	3 customers	—	UNVERIFIED

The customer count matched. The churn claim returned UNVERIFIED because PreDealCheck refuses snapshot churn. Period churn requires cohort data. The engine reported the limit honestly instead of guessing.