Whale Check

Some businesses are naturally whale-shaped. B2B accounts sell a few large contracts a quarter. Furniture retailers ship a small number of high-ticket orders alongside many accessory orders. Luxury or wholesale catalogs have order distributions that look like power laws. The pattern itself is not a defect. It becomes a problem because Target ROAS and value-based bidding average their feedback across all the orders Google Ads sees. When a tiny slice of orders carries most of the value, the bidder is functionally optimizing against that slice. If those whale orders stop being tracked properly (a B2B sales-assisted close that does not import, a high-ticket order processed in a separate checkout, an enterprise deal that converts via a different lead form), the value signal collapses while the volume signal stays unchanged. The dashboard looks the same. The bid strategy quietly starts losing money. The second risk is that whales often live in a different tracking path than the long tail. Enterprise leads route through a separate CRM. Wholesale orders use a different payment processor. Each of those paths is its own potential failure mode that the regular conversion audit does not exercise.

1. Confirm whether the whale distribution reflects the real business model or whether one outlier transaction is skewing the data (a test order at $50,000, for example).
2. Document the tracking path each whale segment uses (CRM import, separate processor, enterprise checkout) and confirm each path is independently monitored.
3. Add alerting on the high-value segment specifically. A 24-hour drop in whale volume should page someone.
4. Consider a conversion value rule or value bucket that smooths extreme outliers if Target ROAS is producing volatile bids.
5. Reconcile whale revenue against the commerce or CRM backend monthly.