Fintech case study: CAC reduced 35% in eight weeks through ICP segmentation, paid account restructure, and landing-page routing — without changing creative or ad spend.
Challenge
Customer acquisition cost had crept up over the preceding two quarters from a workable number to one that broke unit economics on the lower-priced product. The growth team had cycled through three ad agencies, refreshed creative twice, and was about to commission a fourth round of campaigns before bringing in senior help.
Approach
Two weeks of diagnosis showed the problem wasn't creative or audience targeting — it was that the paid account treated three structurally different user segments as one. Re-segmentation by intent, a clean rebuild of the Google and Meta accounts around those segments, and a landing-page routing layer that matched each segment to a different funnel produced the bulk of the improvement. Creative was barely touched.
Outcome
CAC dropped 35% in eight weeks. Paid spend stayed roughly flat. The team kept the rebuilt account structure and the routing logic; the agency relationship was renegotiated around managing the structure rather than buying more creative cycles.
Customer acquisition cost
-35%
Eight weeks, with paid spend held roughly flat
Conversion rate (paid → activated)
+28%
Mostly from landing-page routing, not creative
Number of active campaigns
11 → 6
Sharper segmentation, less spend dilution
Hours per week on creative iteration
-60%
Estimated, vs the previous quarter
A Series-A consumer fintech was a year past PMF, with steady organic growth and a paid acquisition programme that had outgrown the team’s ability to instrument it cleanly. CAC had crept from £18 to £31 over six months, which made the lower-priced of the company’s two products structurally unprofitable. The CFO was four weeks from cutting paid spend by half.
Three agencies had been through the account in eighteen months. Each had refreshed creative, run an audience test, and produced a deck that explained why the next quarter would be better. None of them had touched account structure.
The diagnosis was uncomfortable for the team: the paid account had been set up correctly in 2023 for a single product and a single user shape. By 2026, the company served three distinguishable segments — call them new-to-investing, switching from a competitor, and high-balance professionals — through one product line that branched late in the funnel.
Paid campaigns lumped all three together. Creative variants targeted them differently in copy, but the account-level optimisation was averaging across them. The algorithm was learning to find users who clicked, not users who fit the unit economics of the segment they belonged to.
Three structural changes carried most of the load.
Account split by segment. Google and Meta accounts were rebuilt as three parallel sub-accounts, one per segment, with conversion goals tied to the LTV of that segment. The platforms learned three distinct user shapes instead of one blurred average.
Landing-page routing. A small routing layer at the top of every paid landing page asked one question and routed users into segment-specific funnels. The funnels were already there — the team had built them for organic — they just weren’t reachable from paid.
Decision rules instead of monthly reviews. Each segment got explicit thresholds for “scale”, “stop”, and “investigate”. The growth meeting stopped being a creative review and became a fifteen-minute structured run through the rules.
Creative was deliberately untouched in the first six weeks. The team had been blaming creative for two quarters and we wanted clean attribution for the structural changes.
CAC dropped 35% in the eight weeks of the engagement, with paid spend held roughly flat. Conversion from paid click to activated user moved up 28%, almost entirely from the routing layer. The number of active campaigns came down from eleven to six — sharper segmentation, less spend dilution, more learning per dollar.
The agency relationship was renegotiated. The agency now manages the segment-aware structure and reports against the rules, instead of producing fresh creative on a monthly cadence. The team kept ownership of the account structure and the routing logic.
A senior operator who could read a paid account without the agency history attached, a CFO willing to give the structural rebuild eight weeks before pulling spend, and a head of growth secure enough to admit the problem wasn’t creative. Total scope was 2–4 weeks of dedicated time across the eight calendar weeks, costed as a fixed-scope sprint.
"We had been throwing creative at the problem. The actual fix was structural — and it took an outside operator who could see the account without the history."
Start with a 20-minute growth diagnosis. You will know whether the right next step is a sprint, advisory rhythm, fractional support, or no engagement at all.
Last updated: 11 May 2026