CRO fundamentals: where to start when everything needs fixing
When your site isn't converting, it's tempting to change everything at once. A methodical way to find what's costing you sales and fix it in the right order.
Conversion rate optimisation: more orders and enquiries from the traffic you already have.
Most sites do not have a traffic problem. They have a leak. We find where visitors give up, fix it, and prove the fix with data rather than opinion.
skerrin / Fell Rolltop 22L
Fell Rolltop 22L
£185
4.8 from 214 reviewsMoved up
A roll-top daypack in 18oz waxed cotton with bridle-leather straps. Rolls down for a commute, up for a wet day on the tops.
Add to bagFree UK delivery, 60-day returnsNew
- Control A
- 1.1%
- 44 orders from 4,000
- Variant B
- 1.7%
- 68 orders from 4,000
- B against A
- +55%
- 99% confident B is better. Ship it.
Demo data on our Skerrin demo shop. Confidence is a one-sided two-proportion test on the counts shown.
Who this is for
The mechanics are the same whether the goal is an order or a phone call. The details are not, so we treat them separately.
E-commerce brands
Shopify, WooCommerce, Magento or headless. Anything with a cart and a checkout.
- Product page and collection page tests
- Checkout friction, shipping and payment fixes
- Cart abandonment and post-purchase flows
- Mobile-first layout and speed work
Local service businesses
Trades, installers, clinics, agencies. Anyone whose website exists to make the phone ring.
- Click-to-call and enquiry form placement
- Quote request forms that people actually finish
- Trust signals: reviews, accreditations, guarantees, photos of real work
- Area pages and service pages built to convert, not just rank
What we actually do
- Measurement first
- Before we change anything we make sure GA4, call tracking and form tracking are recording every conversion. You cannot optimise what you cannot see.
- Find the leak
- Funnel analysis, heatmaps, session recordings and short user tests show us where people hesitate, scroll past, or leave. We rank the problems by revenue impact, not by how easy they are.
- Test, do not guess
- Where traffic allows we run A/B tests and only ship winners. Where it does not, we use before-and-after measurement with clear success criteria agreed up front.
- Fix the whole path
- Ad copy, landing page, form, follow-up email or call script. A conversion is a chain, and we work on the weakest link even if it is off the website.
- Speed and mobile
- Most traffic is on a phone, often on a poor connection. We treat page speed and mobile layout as conversion work, because they are.
How an engagement runs
Audit
Two weeks. Analytics review, a map of where people drop out, an expert walk-through and recordings of real visits. You get a prioritised list with expected impact.
Quick wins
Fixes that need no testing: broken tracking, missing trust signals, form fields nobody needs, slow pages.
Test programme
A rolling roadmap of tests, each with what we expect, how we'll measure it and when we'll call it. Typically two to four live at once.
Report and repeat
Monthly results in plain numbers: conversion rate, revenue or leads per visitor, and what we learned. Winners stay, losers teach us.
What you get
- Conversion audit with prioritised fixes
- Tracking setup and verification (GA4, call tracking, form tracking)
- A test plan and a log of what each test taught us
- Designed and built test variants
- Monthly performance report
- Handover documentation for your team
- Conversion rate, DTC brand, over 6 years
- 0.8% → 4.7%
- Conversion rate, fashion retailer
- 1.2% → 3.7%
- Annual revenue, same DTC brand
- £287K → £10.3M
Clients stay anonymous. The figures come from their own platform reports, and we're happy to walk you through them.
Why small gains add up
CRO pays back because it multiplies everything else. Every pound you already spend on ads, SEO and content earns more once the site converts better, and the gain repeats every month without extra spend.
A store with 10,000 visits a month at 1.0% conversion and £60 average order makes 100 orders, £6,000. Lift conversion to 1.5% and that is 150 orders, £9,000. Same traffic, £3,000 more each month, £36,000 a year.
A trades website with 800 visits a month at 3% enquiry rate produces 24 enquiries. At 5% it produces 40. If four in ten become jobs at £250, the extra 16 enquiries are worth around £1,600 a month before any repeat work.
Because the lift is permanent, the question is not whether the fee is covered in month one, but how many months the improvement keeps paying. In our experience that is years.
Worked examples are illustrative. We build the same calculation from your real numbers before you commit, so you can see what a realistic lift would be worth to you.
Further reading
Five email flows, and when to use them
Forget vanity metrics. The automated sequences that do the selling, how to set them up, and the ways they quietly go wrong.
Product page optimisation: beyond the basics
You've got decent photos and written descriptions. Now what? Techniques for product pages that turn browsers into buyers.
Questions we get asked
How much traffic do I need for CRO to work?
For statistically valid A/B tests you need roughly a thousand conversions a month per test. Below that we still improve conversion rates, but we rely on analytics, user research and before-and-after measurement instead of split tests. Most local service websites fall into the second group and still see clear gains.
Does CRO work for a lead-generation website, not just a shop?
Yes. The conversion is a phone call or a form instead of an order, and the same principles apply: clear offer, obvious next step, proof you are trustworthy, and no friction. We track calls and forms per channel so you can see cost per lead, not just traffic.
How long before I see results?
Quick wins from the audit usually land inside the first month. A test programme shows compounding results over three to six months. We agree the success metric with you before we start.
Will you need access to my site?
We need analytics access and either a test copy of your site or a testing tool. For Shopify and WordPress we can make changes ourselves. For custom platforms we work with your developer and hand over specs.
How do you decide whether CRO is worth it for us?
We look at your traffic, current conversion rate and the value of an order or a job, then estimate what a realistic improvement would be worth per month. If that number does not comfortably exceed the cost of the work, we will say so and suggest something cheaper, usually fixing tracking or a single landing page.
Show us your site. We'll show you the leaks.
Send us your URL and we will tell you the first three things we would fix, free, before you commit to anything.
Tell us what you need