Roche calls it the case for subtraction. A big part of conversion rate optimisation is deciding what not to show, and who not to show it to. For UK SMEs staring at IRP Commerce's 2.23% average session conversion (August 2026) and a packed Black Friday calendar, that framing beats another Monday trade-meeting add-on.
Why shops keep adding noise
Every ecommerce team knows the Monday call when sales are soft and leadership wants something new on the site. The default answer is addition: a banner, a trust strip, a timer, a free-delivery shout.
Roche's point is structural, not lazy. Additions are visible. You can point at them in a deck. Subtraction has no dashboard. Restraint rarely gets credit. He points to Leidy Klotz's research in Subtract: people systematically overlook subtractive solutions, especially under cognitive load — exactly the state of a trade meeting with soft numbers and a dozen other fires.
Charley Bader of Made With Intent put the operational version cleanly: most brands already have the right features on the site; the job is giving them to customers when they need them, not as a blanket rule for everyone.
What Cotswold actually changed
Cotswold Outdoor Group runs Cotswold Outdoor, Snow+Rock and Runners Need. Outdoor retail is promotion-heavy, and for years each offer closed with a site-wide countdown timer — same urgency whether you were browsing jeans or footwear.
The timer had "won" an A/B test, so it stayed for everyone. An average win says nothing about who it hurt. Roche rebuilt it in three moves:
- Off site-wide — timer only on product pages
- Promotion-specific — only on products in the offer
- Intent-gated — an agentic campaign shows it to visitors who look ready to buy
Urgency pushes buyers through checkout. For someone still in discovery mode, the same urgency can drive them away. Cotswold set revenue as the primary metric. Roche described the uplift as "fantastic", with secondary metrics up too. Fewer people saw the campaign; the lift among those who did was larger because it was targeted. The honest caveat: the test changed placement, product scope and audience at once, so it proves the approach, not a clean single-variable effect. Primary metric still won.
A second, earlier-stage example at Runners Need: visitors steered toward in-store gait analysis saw free delivery and payment trust icons stripped from the product page — because those answers questions they are not asking yet. Early reads: more gait-analysis interaction and bookings, without hurting sales from shoppers who did not need that path. Cotswold had already seen a similar pattern cutting returns by showing sizing guidance only to high-intent shoppers at the size-choice moment.
The UK conversion backdrop
Subtraction only makes sense if you know what "good" looks like in your category. IRP Commerce's August 2026 UK/Ireland panel put overall session conversion at 2.23%, up from 1.93% a year earlier (+15.55%). Category spread is huge:
| Category (UK, Aug 2026) | Conversion rate |
|---|---|
| Arts and crafts | 5.81% |
| Health and wellbeing | 3.31% |
| Kitchen and home appliances | 2.98% |
| Pet care | 2.56% |
| Sports and recreation | 2.11% |
| Fashion, clothing and accessories | 1.86% |
| Toys, games and collectables | 1.64% |
| Food and drink | 1.58% |
| Baby and child | 0.57% |
Sports and recreation — Cotswold's neighbourhood — sits near the middle. Fashion still converts under 2%. A site-wide urgency widget that "won" an A/B test can still be taxing the wrong half of that funnel.
Baymard's long-running checkout research still puts average documented cart abandonment around 70%, with unexpected extra costs the top cited reason among shoppers who abandoned for reasons other than browsing (~40%). Forced account creation and long or complex checkouts remain high on the list. Subtractive CRO is not a substitute for showing delivery costs early or offering guest checkout — it is how you stop stacking urgency on top of an already noisy path.
A practical subtractive checklist for UK SMEs this peak
You do not need Cotswold's full intent stack to steal the idea. Start with decisions you can explain in a Monday meeting.
- Pick one always-on element. Banner, timer, sticky offer bar, floating chat, trust strip on every PDP. Ask: who is this for, and what question does it answer?
- Split discovery vs ready-to-buy. If a visitor is still reading size guides or comparing ranges, urgency and "buy now" chrome are often noise. Keep trust and delivery clarity for the checkout-bound cohort.
- Scope promotions to the products in the offer. Site-wide countdown on unrelated categories is the "footwear offer while browsing jeans" problem Roche described.
- Measure revenue (and guardrails), not just relative uplift. Narrow audiences inflate percentage lifts. Protect sales among visitors who never see the element you hid.
- One change at a time when you can. Cotswold's first timer test was deliberately clean (timer or no timer) to prove the concept before modelling a wider "friction hierarchy".
- Fix Baymard basics first. Delivery costs on PDPs, guest checkout, shorter forms, express pay (Apple Pay / Google Pay / PayPal), mobile speed. Subtraction on a broken checkout is theatre.
- Report the thinking, not only the widget. Leadership cannot see something you removed. Roche's playbook: win early, drip-feed the concept into trade meetings, report losers as honestly as winners.
Why this matters for Clarity Growth
Clarity Growth helps UK ecommerce teams improve conversion through connected storefronts, cleaner journeys and honest measurement — including Shopify builds and growth optimisation work. Cotswold's 12 October write-up is a timely reminder that peak season is when sites accumulate the most chrome. IRP's 2.23% average and Baymard's abandonment pattern say most UK stores still leave money in friction, not in missing features. We help clients audit which on-page elements actually serve which intent stage, wire promotion logic to the right products and audiences, and keep revenue and guardrail metrics in the same report — so "less on the page" becomes a commercial decision, not a design preference.
Sources
- The case for subtraction: how Cotswold Outdoor Group grew revenue by showing less — Made With Intent (12 October 2026)
- Ecommerce Market Data and Ecommerce Benchmarks for August 2026 — IRP Commerce (August 2026)
- Ecommerce Conversion Rate: What's Good in the UK? — Delivered Social (11 October 2026)
- Reasons for Cart Abandonment — Baymard Institute
- Ecommerce Checkout UX Guide — Baymard Institute