Case studies
What the dashboard missed
Each case shows what the tools said, what the connected data showed, and the decision that followed.
80% of ad spend was on the product that ends the relationship.
Read the case →1.6xnew customer ROAS, against a 2.0x break-evenThe 3.1x ROAS was mostly subscribers reordering.
Read the case →2xrepeat rate of full-price buyers vs code buyersThe 20% welcome code was training discount buyers.
Read the case →3xmore likely to cancel after box one on the smallest boxThe small box looked cheaper to acquire. It was the most expensive customer.
Read the case →+38%contribution per order when the bundle was the first purchaseOne product did the acquiring. The bundle did the profit.
Read the case →41%of new Amazon buyers came from regions where DTC already soldAmazon growth was mostly DTC customers switching channels.
Read the case →-$4.10contribution per TikTok Shop order after fees, commissions and adsTikTok Shop added $40K a month in sales. It also lost money on every order.
Read the case →31%of Meta-driven sessions landed on a product that was out of stock or low on variantsMeta did not break. The best sellers were out of stock.
Read the case →17%of new customers named the creator campaign in the post-purchase surveyThe creator campaign showed zero ROAS. It brought in one new customer in six.
Read the case →Cases marked illustrative are composites built from patterns we see at this stage, with details changed. Named client results are added as they are cleared.