Conversion rate optimisation
Most "we should redesign the site" conversations are really someone’s hunch about what’s wrong, argued confidently enough to sound like a decision. CRO is the discipline of testing that hunch against real visitor behaviour before you spend the budget on a full rebuild.
We’ll look at your funnel data and show you where visitors are actually dropping off, with no guesswork.
A senior CRO strategist will reply within one working day.
Someone in the room always has a theory about why the site doesn’t convert: the hero image, the button colour, the headline. CRO exists because most of those theories are wrong, and a full redesign built on a wrong theory is an expensive way to find that out.
Our CRO services
Before testing anything, we find out where visitors actually drop off: funnel analysis, heatmaps, session recordings and on-site surveys where appropriate. This stage replaces opinion with evidence about where the real problem sits, before anyone proposes a fix.
Research before a single wireframe is drawnWe run structured experiments on real traffic, testing one clear hypothesis at a time and letting the data decide the winner rather than whoever argued most confidently in the meeting. Each test builds on the last, so the queue compounds rather than resets.
A test queue, not a single betNot every idea deserves a test slot. We score hypotheses against potential impact, confidence from research, and ease of implementation, so the testing roadmap runs the ideas most likely to move the number first, not whichever idea is loudest in the room.
A prioritised list, not a guessing contestQuantitative data tells you where people drop off; moderated usability sessions tell you why. We run both, because a heatmap can show a dead click without ever explaining what the visitor was actually trying and failing to do.
Watching real people struggle, not guessing where they wouldPaid traffic landing pages get a tighter testing loop because the cost of each visitor is known and the sample sizes needed to reach significance arrive faster. We treat these pages as a continuous experiment, not a one-off build signed off and left alone.
Faster page loads, tested the same way as everything elseCRO compounds over time: each test teaches something that informs the next one. Our ongoing programmes keep a steady cadence of tests running, with monthly reporting that shows what was learned, not just what shipped.
A test roadmap that keeps running, not a one-off reportWhat CRO actually is
"Let’s redesign the site" is usually shorthand for a theory nobody has actually tested: the homepage is too cluttered, the CTA isn’t clear enough, the checkout has too many steps. The theory might be right. It also might not be, and a full redesign commits the entire budget to finding out either way, all at once, with no way to isolate which change actually helped.
CRO inverts that: pick the single highest-confidence theory, build the smallest version of the change that tests it properly, and run it against a control on real traffic until the result is statistically meaningful. If it wins, roll it out. If it doesn’t, you’ve lost a fraction of what a full redesign would have cost, and you know something true about your visitors that you didn’t know before.
Done consistently, this turns "redesign" from a single expensive guess into a compounding series of small, evidenced wins — and it tells you, honestly, when a bigger structural change really is warranted rather than just assumed.
See it in the test
Not a redesign on a hunch — a controlled test with a real winner, run until the numbers are big enough to trust.
Illustrative A/B test mockup in wtfont’s own design language — not a real client test.
Our process
Every stage produces something you can check: data, a documented hypothesis, a test result with the numbers shown, not just an opinion dressed up as a conclusion.
We study funnel data, heatmaps, session recordings and on-site feedback to find where visitors actually drop off, before proposing a single fix.
Research findings become specific, testable hypotheses, scored by likely impact and confidence, not just listed as ideas.
Each test isolates one variable cleanly, with a clear success metric agreed before the test goes live, not interpreted afterwards.
Tests run until statistical significance is reached, monitored throughout rather than checked once and left alone.
Winning changes roll out site-wide, the result feeds the next hypothesis, and the testing queue keeps moving.
Who it’s for
That’s a theory, not a plan. We help you test the theory on a slice of traffic before committing the full redesign budget to it.
More traffic on a leaky funnel just means more leaked visitors. We find where the leak actually is before recommending anything.
A redesign without a tested hypothesis behind it is a coin flip with a big budget. We make sure the next change is evidenced, not another flip.
Good design and high-converting design overlap often, but not always. We tell you honestly when the data disagrees with the opinion in the room.
Why hire a CRO agency
Working together
We nearly spent the whole budget on a redesign that testing proved wouldn’t have worked.
Finally a team that tells us when a test didn’t win, instead of hiding it.
Small tests, stacked up over a year, moved the number more than any redesign we’d tried.
Sectors
Same research-before-testing discipline, tuned to what actually drives the decision in your sector.
Product page and checkout testing focused on add-to-cart rate and completed purchase, with mobile and desktop tracked separately.
Trial sign-up and onboarding flow testing, where a small conversion improvement compounds significantly across the customer lifetime.
Trust and clarity testing on application and quote flows, where hesitation is often about risk perception rather than design.
Booking flow testing across date selection, pricing display and payment, where abandonment often spikes at a specific, findable step.
Enquiry and application form testing, where reducing friction at the right step measurably lifts completed applications.
Form length, CTA clarity and landing page testing focused on lead quality as well as volume, not volume alone.
Tools and tech
Reliable testing data beats a flashy dashboard nobody checks. These are the platforms our CRO team works in.
What you get
No "trust us" redesign and no vague promise of improvement. Here is what lands in your inbox and your calendar.
And always: full visibility into every test running, including access to the raw data, not just a summary slide.
Beyond CRO
For when research and testing point clearly toward a genuine structural rebuild.
Product and checkout rebuilds informed by what testing already proved mattered.
More qualified traffic arriving at a funnel that is already proven to convert it well.
Content built to feed a funnel that has already been tested and tightened.
Ad spend that performs better once the landing experience it funds has been proven to convert.
The measurement foundation every test and every hypothesis in this programme depends on.
Who you’re working with
Every CRO programme at wtfont. has one named strategist who runs it end to end: they lead the research, prioritise the hypothesis roadmap, and are the person who explains a result to you honestly, win or not. No handoffs to a team you have never met.
Swap this block for the real person: their name, photo, years of testing and research experience and a programme they are proud of. Showing who actually designed the tests is one of the easiest trust signals to add for anyone checking your credibility before committing.
The bigger picture
Acquisition costs have risen steadily across paid search, paid social and even organic visibility, as AI-generated summaries absorb more of the clicks that used to reach a website directly. Every visitor who does make it through now carries more acquisition cost behind them than they did two or three years ago, which raises the value of converting that same visitor rather than simply chasing more of them.
At the same time, the tools for running rigorous tests have become more accessible, removing the old excuse that testing required a dedicated data team and six-figure tooling budget. A mid-sized site can now run a credible testing programme with the same statistical rigour that was once reserved for large ecommerce platforms, which makes skipping that step harder to justify than it used to be.
The businesses still treating "redesign" as a single confident bet, rather than a sequence of tested decisions, are increasingly the ones left wondering why the new site performs no better than the old one. The ones treating it as an ongoing experiment compound small wins quarter after quarter, and arrive at any eventual bigger rebuild with far more certainty about what actually needs to change.
None of this requires abandoning design instinct. It requires treating that instinct as the source of hypotheses, not as the final word on what gets built.
CRO guide
Straight answers to the questions every business asks before committing to a testing programme.
It depends on your current conversion rate and the size of effect you're trying to detect, but as a rough guide, a few thousand monthly visitors to the page being tested is usually enough to reach significance within a reasonable timeframe. We'll calculate the realistic timeline for your actual traffic before starting.
That's a valid and useful result. It tells you the change didn't matter enough to pursue, which saves you from building it site-wide on a hunch. We report inconclusive tests honestly rather than stretching the data to find a story.
Test first whenever you can. A full redesign makes sense when research already points clearly at a structural problem testing alone can't fix, such as an outdated information architecture, not as the default first move.
Early research findings often surface within the first two to three weeks. The first completed test typically reports within four to eight weeks depending on traffic volume, with the programme compounding from there.
FAQ
Research into where visitors drop off, a prioritised hypothesis roadmap, A/B and multivariate testing run to statistical significance, and honest reporting on every test, win or not.
Monthly retainer based on testing cadence and traffic volume, with an initial research phase scoped separately. No long contracts; you can pause or stop with notice.
Not usually. Most tests run through a testing platform without needing engineering resource, though more complex tests involving backend logic may need light developer input, which we'll flag upfront.
Platforms like VWO or Optimizely for running tests, alongside Hotjar or FullStory for qualitative research and GA4 for underlying analytics, chosen to fit your existing stack where possible.
Yes, regularly. CRO sits alongside acquisition work rather than replacing it, and often improves the return on whatever acquisition spend is already happening.
It varies widely by starting point and sector, so we avoid quoting a blanket number upfront. We'll give you a realistic range once we've seen your actual funnel data in the research phase.
Both, prioritised by likely impact relative to effort. Small changes with strong research backing often get tested before bigger, riskier ones, because they reach significance faster and build the evidence base.
No. Testing platforms are built to add minimal page weight, and we monitor site speed throughout any programme to make sure testing itself never becomes the thing hurting conversion.
This page is specifically about ongoing testing and experimentation on a site that already exists. Our web design service covers new builds and full redesigns, which we'll always recommend testing into rather than guessing at where possible.
From the blog
Guides and playbooks from the team. Placeholder titles, link them to your posts.
Sample size, significance and the mistakes that invalidate results.
What the data says about common redesign assumptions.
A practical guide to choosing the right research method.
Scoring hypotheses by impact, confidence and effort.
Let’s talk
30 minutes with a senior CRO strategist. No pitch deck, no obligation. Hang up anytime.
A senior strategist will reply within one working day.
Trusted with the data
Teams who stopped guessing
and started testing
From single landing pages to full-funnel testing programmes, we’ve helped teams replace confident-sounding opinions with evidence they could actually act on.
- Senior CRO strategists on every programme, not junior handovers
- Full access to raw test data, not just a summary report
- Honest reporting that includes the tests that didn’t win
Talk to our CRO teamSome of the brands we test for
Replace placeholders with real client logos