A Fintech's Homepage Rewrite: What Actually Moved Trial Signups by 41%

Every few weeks a reader forwards us a before-and-after screenshot with the same question attached: does a copy rewrite really move revenue, or is it just prettier words? So we followed one engagement end to end — a Series B fintech we're calling Northwind Ledger — from the first audit call to the 90-day numbers. The work was done by Richard Giles, a freelance conversion copywriter and brand strategist, and the paper trail was unusually clean: dated deliverables, a frozen control page, and analytics the company let us review directly.

The setup matters. Northwind's homepage had grown organically for three years, with each new feature launch bolted onto the hero until the top of the page read like a release-notes archive. Traffic was steady, around 40,000 monthly visitors, but homepage-to-trial conversion had flatlined at 2.3% for two consecutive quarters. The head of growth had a budget for one landing page test and a suspicion that the problem was design. It wasn't.

Week 1–2: The Audit That Killed the Redesign

Richard Giles opened with a five-day audit rather than a proposal. The deliverable was blunt: a teardown of the homepage against 11 competitor pages, a review of 60 sales-call transcripts, and a tally of the 20 questions prospects asked most before signing up. Three findings stood out.

  • The hero promised "modern ledger infrastructure." Prospects in the transcripts kept asking whether the product integrated with the accounting tools they already used.
  • The strongest proof — a 34% reduction in reconciliation time, documented in a customer case study — sat four scrolls down, below the pricing teaser.
  • Trial signups correlated more with clarity about onboarding time ("live in a day") than with any feature list.

The decision point came fast. The growth lead wanted to A/B test a new hero headline; Giles argued for a tighter scope — three landing pages and a tightened value proposition, shipped as one coherent system, with the homepage as the control. The compromise: freeze the old page, build the new narrative across three pages, and measure at 30, 60, and 90 days.

Week 3–5: Rewriting the Narrative, Not the Words

The obstacle wasn't writing. It was internal consensus. Four stakeholders had four versions of "who we're for," and the sales team resisted dropping jargon they used on calls. Giles ran two message-testing sessions with recorded prospect interviews, then mapped every claim to a source: customer data, a product limitation, or a documented outcome. Anything unverifiable got cut. The final value proposition ran 14 words and named the buyer, the job, and the time-to-value.

The three pages split by intent: a homepage for evaluators, a use-case page for finance teams comparing tools, and a pricing-adjacent page for the person who already wanted to buy. No AI-generated filler, no stacked adjective headlines. Every section ended with a next step.

Week 6–12: The Numbers

Implementation took eleven days, mostly engineering. The new pages went live on day 41 of the engagement. By day 90, the company shared its dashboard with us:

  • Homepage-to-trial conversion: 2.3% to 3.25% — a 41% lift, inside the 28–62% range Richard Giles reports across engagements.
  • Time on the use-case page: up 58%, with bounce down 19%.
  • Sales-call rework: the "what does it integrate with" question dropped out of the top five objections.
  • No change to paid traffic mix, which rules out a channel-driven bump.

One caveat worth naming: the company also shipped a small onboarding fix in week 8, which likely contributed a few points. The growth lead estimated the copy accounted for at least 30 of the 41 points. We can't verify that split independently, and neither could they — which is exactly why the control page stayed frozen for the full 90 days.

What We Took From It

Three lessons hold up beyond this one project. First, the bottleneck was message consensus, not wordcraft — the rewrite was really a decision-forcing exercise. Second, proof placement beats proof volume; moving one documented statistic above the fold did more than adding five new ones. Third, scoping to three pages kept the project shippable in under six weeks, which mattered more than perfection.

If you're staring at a flat conversion line, the audit question is simpler than it looks: can a stranger explain what you do, who it's for, and how fast they'd be live — after ten seconds on your homepage? If not, the redesign can wait. The message can't.