Reviews

Not a wall of five stars. A few people who used the work.

Helen’s payback-clock module in Curve to Cash is the one I still open. We had been reporting LTV to the board and payback to the UA stand-up as if they were the same clock.

Naomi Ellis, finance partner, Edinburgh

Signal Notes was cheaper than I expected and thinner on Android. The trial-leakage table was enough to change how we cut a GB promo week.

Owen, Cardiff

★★★★☆

Forecast Desk hours were sharp on hybrid SKUs. They would not look at our MMP dashboard, which I found stubborn at the time and later agreed with.

Rated on our internal vendor note

A client in fitness subscriptions: we stopped treating restorative purchases as new LTV after the naming-the-unit session.

I joined from a marketplace app. The observed-curve homework exposed that our “retention” was mostly repeat sellers, not buyers. Uncomfortable, useful.

Rina K., operator, Birmingham

Evenings in UK time worked; the recording for week six lagged two days, which clashed with our board prep. Content of the tails lecture was still the right kind of cautious.

First name withheld · city: York

Case: grocery delivery, GB-heavy mix

A delivery app arrived with a single LTV used for both London grocery and a smaller convenience SKU. Curve to Cash week two forced a split. The convenience cohort paid back media in eleven days and then flattened; grocery took thirty-eight days and kept a thin tail of repeat baskets. UA had been bidding as if both were grocery.

They did not “win” a higher LTV. They stopped overpaying for the SKU that looked busy on blended ARPDAU. The limitation they now print: promotional free-delivery weeks are excluded from the base curve, which still makes finance nervous in Q4.

Case: hybrid IAP inside a language app

The team sold a yearly subscription and a pack of “coach credits”. Their LTV summed both at face value. The hybrid module showed credits were often bought by already-subscribed power users. After the correction, the yearly plan still carried the forecast; credits became a margin line, not a second lifetime.

Forecast Desk later reviewed their board pack. We would not bless a D540 number built from fourteen weeks of data. They published D180 with a band instead, and the round proceeded on that more modest claim.