# Case Study: Three Separate Failures, None Visible From Inside the Advertising Platform | Crank Source: https://wearecrank.com/case-studies/beauty-skincare-operational-audit-2024 A gift-set range that couldn't be advertised, a checkout giving no reason to trust it, and budget flowing to the wrong campaign — none of it visible from the platform's own reporting. --- Beauty & Skincare — 2024 Audit # Three Separate Failures, None Visible From Inside the Advertising Platform A gift-set range that couldn't be advertised, a checkout giving no reason to trust it, and budget flowing to the wrong campaign — none of it visible from the platform's own reporting. Challenge An entire gift-set range was ineligible for shopping campaigns because of missing product identifiers, the checkout gave no reason to trust it over a named competitor's, and a genuinely better-converting campaign was being starved of budget while a worse one ran freely — three separate problems, none visible from inside the advertising platform. Approach Fix what was structurally broken first — the feed, the conversion events, the exclusions between new and existing customers — then reallocate budget to what the account's own data already showed was working, and act on the platform's own quality signals on creative. Outcome Coverage was restored, a starved campaign acquiring customers at £13.94 was funded properly against one running freely at nearly double the cost, and the account had 303 additional monthly sales of modelled headroom at £16.02 each once the daily budget stopped switching off the best-performing days. 01 ## Where the money was already lost before the campaign ran All gift sets were missing their product identifiers, making the entire range ineligible to appear in shopping or automated campaigns — nothing in the advertising platform indicated a problem. Once found, identifiers were added and the range became eligible for the first time. The checkout itself gave shoppers little reason to choose it. Accepted payment methods were shown nowhere on the site, and no instalment option was offered while a direct named competitor provided one. Hero banners ran taller than a standard screen with very small calls to action, so the primary action on the page wasn't visible without scrolling — and that same named competitor matched the pricing while offering free delivery above a £30 threshold, which this brand did not. 02 ## Measurement rebuilt, then acted on The advertising platform's conversion interface was only partially implemented — basket abandonments, searches, registrations and purchases weren't all being captured. A full event set was implemented, making basket abandonment retargeting possible at all and letting high-intent visitors be shown the products they'd left behind. New and existing customers had also been served identical advertising with no exclusions, making it impossible to separate the cost of winning a customer from the cost of keeping one. Once separated, retention could be run at its own economics rather than hidden inside a blended figure. 03 ## Budget going to the wrong campaigns, on purpose Daily budgets were being exhausted before the end of the highest-converting days, switching off the best-performing activity precisely when it was working. Separately, one campaign was acquiring customers at £13.94 while another in the same account cost £26.06 — the cheaper one constrained by budget, the dearer one not. Once reallocated, the account had headroom modelled at 303 additional conversions a month at £16.02 each, on an 8.6 per cent conversion rate — demand the budget simply hadn't been able to reach. 04 ## Creative and audience the account already had evidence for Several adverts were rated poor for strength by the advertising platform's own quality measure, limiting how often they were shown — a signal the platform provides for free and nobody had acted on. Separately, analysis showed one in-market interest segment converted at the highest rate, and that 70 per cent of site visitors used a single device type — neither of which was reflected in the targeting running at the time. ## What was found, once someone looked past the reported numbers £13.94 vs £26.06 Cost per acquisition, starved vs. funded campaign 303 @ £16.02 Modelled additional monthly sales, once unconstrained Months A gift-set range sat unadvertisable before identifiers were fixed 70% Visitors on one device type, unreflected in targeting Drawn from Crank's account audit conducted in 2024 for this beauty and skincare retailer. Client not named. ## Want to know what this looks like in your own account? These findings came from asking straightforward questions before any budget moved. We're happy to do the same for yours. [ Talk to Wearecrank](https://wa.me/447457416922?text=Help%20my%20business)[Get in touch](/contact) [← Back to all case studies](/case-studies)