Revenue was up. The business was getting worse.

A DTC supplement brand built around a subscription model and a genuine belief in what they were selling. Growing by the numbers. The cohort data told a completely different story - and almost nobody was looking at the cohort data.

Retention was eroding. New customers were signing up, making it look like growth, then quietly cancelling before they became profitable. Acquisition and retention were being treated as the same problem, managed by the same people with the same KPIs. Neither was being done well as a result.

Stop the leak. Make acquisition profitable on real cohort economics. Build a retention programme that didn't rely on discounting.

Split the business into two separate P&Ls with different success metrics. Rebuilt the post-purchase experience - sequences built around habit formation and product education, not discount codes. Rebuilt paid acquisition to target buyers with genuine subscription intent signals rather than one-time purchase patterns.

• Month-3 retention improved from 41% to 63% across the next two subscriber cohorts
• 6-month LTV up 34%
• Blended CAC down 19% as paid stopped acquiring customers who churned immediately
• Business hit payback period 6 weeks faster than the previous model projected

"The hardest part was admitting that the growth number we were celebrating was masking how bad the underlying business had gotten. Once we saw it clearly, fixing it was the easy part."- CEO