By the tool you already run

Four defensible readings of one table — real figures from What the raw table costs you, laid out as Excel would present them. A rendering, not a screenshot.

revenue_treatments.xlsx
D5=MAX(C2:C5)-MIN(C2:C5)
ABCD
1treatmentlinesrevenue USDvs naive raw USD
2raw, naive q x p1,067,37119,287,250.570
3raw, cancellations removed1,047,87720,813,918.431,526,667.86
4shared load_clean()1,024,82620,050,680.95763,430.38
5load_clean() + sheet dedup1,003,21419,642,692.15355,441.58

The spread in D5 is $2,366,335 — the gap between the highest and lowest defensible reading of the same table.

By practice

What each engagement actually hands over. Three of these are not things a reporting tool produces at all — a data contract, warehouse documentation and a model card come out of the work itself.

An accuracy pack scored against every baseline — real figures from Weekly SKU demand forecasting. A rendering of the deliverable, not a screenshot.

Forecast accuracy pack13-week holdout
MethodAggregate WMAPEPer-SKU median WMAPETotal biasWMAPE vs best baseline
Naive (last week)0.82780.8809-18.8%-14.4%
Moving avg (4wk)0.72390.7915-27.9%+0.0%
Seasonal naive (52)1.171.24+46.6%-61.0%
Train mean0.75820.8332-13.7%-4.7%
Ridge0.80581.00+1.6%-11.3%
GBM (log target)0.68240.7762-22.4%+5.7%
GBM (mean loss)0.92341.02+32.6%-27.6%
GBM (median loss)0.68480.7761-26.6%+5.4%

Every baseline the model was scored against, not just the one it beat. A pack that shows only the winning comparison is a sales document.