Barking Cat, Commerce AdvisoryBarking Cat Advisory Book a callBook

The Commercial Experience Index · Edition Three  /  Outerwear and Function  /  November 2026

Help first. Ask later.

Twenty outerwear and function brands, scored out of 100 on the experience each chose to publish. Read from the live UK shopfront, not the balance sheet. Arc’teryx score 72 and top the table. The four highest-scoring sites never put a sign-up pop-up between the shopper and the product.

DOWNLOAD THE FULL REPORT (PDF) 24 pages · 280 KB · No form, no email, no gate

The season for waterproofs

The jacket is the easy part.

November is when the category earns its year. The weather turns, Black Friday arrives, and twenty brands built on protecting people from the elements compete for the same customer, often for the same jacket, in the same week.

The Commercial Experience Index asks a narrow question of each of them: does the experience the brand has built help that customer move from arrival to commitment. It does not see the profit and loss, the traffic, the conversion rate or the margin. It sees what a customer meets on the site, and what a search engine or AI assistant can read from it, and it scores that.

Every brand was walked twice on 17 November 2026: once by the author as a shopper, and once against a fixed ten-point read standard covering pop-ups, navigation, filters, fit information, product recommendations, express payment and the checkout gate. Where the two walks disagreed, the site was checked again, and the shopper’s first-hand read stood where the evidence held. Every score traces to something a reader could observe or re-run.

Why these twenty

Blending fashion with function.

The remit was functional outerwear, read in the weeks before the worst of the winter arrives, when a waterproof stops being a wardrobe choice and becomes a necessity. The twenty span the ways the category now sells: technical mountain specialists, sailing names with decades of heritage, direct-to-consumer challengers, family and value brands, and the labels that have carried performance outerwear from the hill onto the high street. Each sells a jacket that has to survive bad weather and look right in town, and that blend of fashion and function is what makes the category a sharp test of commercial experience. The customer is buying a look and a specification at the same time, and the store has to sell both.

The headline

The four highest-scoring sites in the set never put a sign-up pop-up between the shopper and the product. Fifteen of the other sixteen did: on arrival, on the listing page, or on the product page itself, at the moment the customer was reading the detail they came for. Two asked again after being told no.

Arc’teryx score 72 and Patagonia 70, with Alpkit and The North Face level on 67. At the other end sit Paramo and Rohan on 53, with Musto and Craghoppers on 55.

The table breaks in two. Thirteen brands score 62 or above and seven score 57 or below, with nothing in between. Every brand below the break carries at least one fault a first-time shopper meets before payment: a size guide that ends the journey, recommendations aimed at the wrong customer, badges that contradict each other, filters that will not open, or no express wallet at any point. The full ranked table, with every measure, is below.

The table

All twenty brands, ranked out of 100. Each brand is scored across eight measures of commercial execution: Proposition, Experience, Persuasion, Checkout, Copy, Performance, SEO and AI Readiness, each read from the live shopfront and scored 1 to 10. The total out of 100 is the headline.

#Brand Prop Exp Pers Chk Copy Perf SEO AI Total
1Arc’teryx78.16687.77.67.372
2Patagonia87.86686.57.25.570
3=Alpkit67.57577.67.26.267
3=The North Face67.86756.57.67.067
5Rab68.65477.68.08.066
6Gill68.05567.57.68.065
7=Montane77.85557.87.05.264
7=Henri-Lloyd77.34577.27.27.264
7=Dryrobe67.25666.78.17.264
10Berghaus67.25676.26.86.263
11=Jack Wolfskin57.85657.87.65.062
11=Fjallraven77.75476.57.23.862
11=Mountain Equipment66.55776.86.64.062
14Columbia56.25766.26.21.557
15=Helly Hansen57.25455.38.04.556
15=Regatta47.05456.57.57.056
17=Craghoppers57.85455.57.52.055
17=Musto56.25456.57.64.055
19=Rohan46.35557.16.14.053
19=Paramo46.05466.57.13.753

Measure scores are out of 10. Experience, Performance, SEO and AI Readiness carry objective data and are shown to one decimal place. Brands on the same rounded total share a rank. Scores reflect the live shopfront on 17 November 2026 and will move as brands update their stores.

The eight measures

What each measure tests, where the field clustered, and where it separated. Five measures are an operator’s judgement of what the shopper experiences. Three read objective data: speed on a mobile phone, and how well search engines and AI assistants can read the store.

01

Proposition and Brand Clarity

The measure that taxes every visit above it. The test is a cold one: with the sound off, in three seconds, can a stranger tell what this brand is, who it is for, and why it over the alternative. Most of the twenty pass the first question easily. The separation sits in the third, and specifically in whether the brand carries an identity through its store or simply drops the season’s sale into a template.

Patagonia stood alone at the top. Activism, repair and a guarantee on everything it makes run through the navigation, the hero and the grid; the PowSlayer hero sells a layering system rather than a garment. Four brands reached above average behind it. Arc’teryx carries technical authority through archive storytelling and its Freeride Academy. Montane holds a fast-and-light identity in its type system and product facts. Henri-Lloyd tells a sailing story dating to 1963 through an Originals Vintage archive. Fjallraven’s heritage and fox identity are unmistakable from the first screen.

The floor was Regatta, Rohan and Paramo. Regatta leads with discount codes and a collection of other people’s brands, and the author’s read was that it felt like a follower. Rohan has a real travel proposition in “For Every Journey” and sells almost everything on it at a discount, including tiles badged New and 50% off together. Paramo’s homepage showed the words “No video available.” where its hero should be, and gave a first-time visitor little to say what the brand is about.

The finding that matters: Black Friday flattened the middle of the field. Brands with genuine stories (Berghaus’s repair service, Columbia’s cold-weather wit, Mountain Equipment’s “Built Better”) let a sale banner do the talking in the week that mattered most.

02

Experience and Merchandising

The test is whether a customer can move without being made to think, and whether product is presented and sequenced with commercial logic. This edition read it further than before: one menu opened and one filter applied on every site, every product-page tab opened, the size guide followed to see whether it returns the customer to the product, and the recommendations read at three points after the add to bag.

Rab topped the measure on discovery. Its filters work by outcome (most waterproof, most breathable, very light) alongside how-to guides, an operator-authored idea no platform provides. Arc’teryx and Gill follow: Arc’teryx on editorial sub-category tiles and a fit and layering guide that opens in place, Gill on categories built around sailing conditions and a four-point fit scale under the size buttons.

The separation came after the filter. Helly Hansen and Musto both send the customer to a separate size-chart page with no route back to the product. Gill and Dryrobe recommended the right products in sizes that had sold out. Regatta recommended women’s gloves, beanies and socks against a men’s jacket. Rohan recommended one T-shirt five times. Berghaus loaded nine of its 42 large homepage images blank.

Paramo sat at the floor. Its Filter & Sort would not open in Chrome or Firefox on desktop or Chrome on mobile, and the menu did not behave as a menu. A customer can still reach the men’s range by clicking the top link; beyond that, the store gives them little help to narrow it.

03

Persuasion and Buyer Journey

The flattest measure in the edition, and the one the pop-up decided. The test is whether prompts to buy are present where the customer hesitates, and whether they help or interrupt. Fifteen brands interrupted. An interruption at the point of hesitation costs more than the trigger beside it earns, and the measure reflects that across the whole field.

Alpkit led by doing the opposite. A 123-day Christmas returns window sits directly below the hero, placed where a gift buyer hesitates, and no pop-up stands between the shopper and the product. Arc’teryx, Patagonia and The North Face followed. The North Face’s size reassurance (an exchange free of charge within 30 days if it does not fit) sits where the doubt is, and its membership offer waits until a product has been chosen.

Henri-Lloyd finished last. Its arrival pop-up asks for a name, an email address, a phone number and consent to SMS marketing in return for news of a daily deal, with no discount attached. The gap between Henri-Lloyd and the brands above it is the difference between an interruption that offers the customer nothing and one that offers something real.

The behavioural reason is old and well documented. Psychological reactance, described by Jack Brehm in 1966, is the pushback people feel when their freedom to choose is threatened: an overlay that blocks the product, asks for more than an email, or returns after a no is read as pressure, and the shopper resists the source of it. The pop-ups in this edition that offered the least (no discount, a phone number, a second ask) are the ones most likely to provoke it. Google has also treated intrusive mobile interstitials as a negative signal in search since 2017.

No brand reached the upper band. Several deploy genuine triggers (Dryrobe shows reviews inside checkout, Columbia shows product-view counts) and then spend that goodwill on a sign-up form the customer did not ask for.

04

Checkout and Friction

Scored on the logic that friction at the commit moment destroys revenue at its most valuable point. The read is honest about its limit: it goes as far as the front end allows, through the express-pay options at the product page, basket and checkout, the guest gate and the marketing consent, and stops at the first personal-data field. No account was created and no order placed.

Guest checkout exists at all twenty brands; this measure separated on express payment and consent. The North Face, Mountain Equipment and Columbia led. The North Face offers Apple Pay at the product page, basket and checkout. Mountain Equipment offers Shop Pay and Google Pay in the drawer, the basket and at checkout, leaves the marketing box unticked, and prints a delivery date on every line of the bag.

All eight Shopify stores offered an express wallet, and four of them (Alpkit, Gill, Henri-Lloyd and Montane) pre-ticked the marketing box. Rohan signs the customer up by default with an opt-out link. Patagonia offers Apple Pay in the basket and drops it at checkout.

The floor was the six brands that offered no express wallet at any point walked: Craghoppers, Fjallraven, Helly Hansen, Paramo, Rab and Regatta, with Musto alongside on a checkout that offers none at the gate. Fjallraven adds a sign-in-first gate with guest as a secondary link, and on the author’s walk the email step stalled before it advanced.

05

Performance

Objective

The first of the measures read from objective site behaviour rather than judged. It reads how quickly and cleanly each homepage loads and runs on a mobile device, from Google PageSpeed Insights on 17 November 2026. Mobile is the right form factor: most outerwear traffic arrives on a phone.

The category is slow. No homepage in the set scored above 61 for mobile Performance in the lab, and Craghoppers scored 14. Seven of the twenty passed Core Web Vitals in the field: Alpkit, Arc’teryx, Berghaus, Montane, Paramo, Patagonia and Rab. Montane and Jack Wolfskin top the measure, with Arc’teryx close behind. Helly Hansen and Craghoppers sit at the floor.

Two kinds of failure deserve a specific watch-out the blended score softens. The North Face recorded a field layout shift of 1.0, the worst single reading in the set: the page moves under the customer’s thumb as it loads, beneath an otherwise respectable score. Six other brands also fail on layout stability in the field, Columbia, Craghoppers and Regatta among them. And nine brands fail on responsiveness in the field, with The North Face, Regatta and Helly Hansen all taking more than 300 milliseconds to respond to a tap.

06

SEO Readiness

Objective

A measure of readiness rather than performance. It asks whether a brand has built the structural conditions to be found, read from the front end only: titles and descriptions, heading structure, schema presence, crawlable navigation and clean URLs. It does not measure rankings or traffic, which need tools beyond the shopfront and belong to the full audit.

The least dramatic measure in the edition, because most current platforms do the basics well and the band is tight. Dryrobe, Rab and Helly Hansen top it.

Four brands sit below the pack. Rohan finished last: its robots.txt points search engines to a sitemap on localhost, and its category links are not plain crawlable anchors. Columbia, Mountain Equipment and Berghaus follow, Berghaus with an empty meta description on its homepage, the one line a search result shows a customer before they click.

07

AI Readiness

Objective

Scored lower than organic search today, because AI-assisted discovery is a real but still smaller commercial channel. The measure reads readiness: whether an AI assistant or shopping agent can reach the store, read the product, and price it correctly from data the brand controls. It does not measure whether assistants currently recommend a brand, or whether any brand earns money from them; that needs live prompting and belongs to the full audit.

Rab and Gill top the measure on clean, complete product data and full Agentic Browsing results. Arc’teryx follows with the most complete structured data of the twenty and an explicit welcome for the AI crawlers in its robots file.

The floor is Columbia and Craghoppers. Columbia returned no Google merchant listing for its walked product, a product description in its structured data reading “Mountains optional.”, and an Agentic Browsing result of zero. Craghoppers also scored zero on Agentic Browsing.

Three findings carry weight beyond the scores. The North Face served seven of the eight Google merchant listings checked in US dollars on its UK store, so a machine comparing it against a rival reads the wrong price in the wrong currency. Musto has a product listing that Google accepts at a price of £0. And only two of the twenty publish their shipping terms in a form a machine can read, and only four their returns policy; an agent comparing stores cannot see a free-delivery threshold it is not told about.

08

Copy and Voice

The measure that finishes rather than founds. The test is whether the words do a job, and whether the brand sounds like someone or like a template. This edition opened every collapsed tab on the walked product page before scoring, because the best copy in this category is often hidden there.

Arc’teryx and Patagonia top the measure, each writing copy that does technical and brand work at once: Arc’teryx on the Sidewinder archive (“a design replica of our first freeride shell, now with modern fabrics”), Patagonia on what its gear is made from and what happens to it after. Alpkit is the find of the edition. Its product tabs carry the most distinctive voice in the set once opened, from “Munro-bagging and Wainwright-ticking lunch stops” to a care note about ketchup and hot chocolate.

Berghaus (“You tear it up. We fix it.”), Henri-Lloyd, Fjallraven, Mountain Equipment and Rab each sound recognisably like themselves. The floor is eight brands that write correctly and conventionally, with no voice a reader would recognise with the logo removed. Montane, otherwise one of the most specific sites in the set, carries a homepage line reading “challenges environments”.

The twenty brands

Every brand, in rank order. What the store does well, where it leaks, and the single thing worth watching.

72/ 100

1. Arc’teryx

Prop 7 · Exp 8.1 · Pers 6 · Chk 6 · Perf 7.7 · SEO 7.6 · AI 7.3 · Copy 8

The most complete build in the set, and first place in Edition Three. Arc’teryx works for the shopper and for the search engines and AI assistants that read its store, and scores 6 or better on every measure.

Where it did well. Structure and discovery. Editorial sub-category tiles, a fit and layering guide that opens in place, and an added-to-cart panel that recommends packs and a harness rather than another jacket. Copy that does technical and brand work at once, and the most complete structured data of the twenty.

Room for improvement. The jacket’s own fit sits in the product details far below Add to cart, so the customer reading the size buttons has to go looking. Apple Pay and PayPal are offered in the cart and gone at the email step.

Watch-out. The walked product page showed a two-star average at the top, driven on the author’s read by price comments rather than product faults. For the most expensive shell in the set, the first number a customer reads is a reason to hesitate.

70/ 100

2. Patagonia

Prop 8 · Exp 7.8 · Pers 6 · Chk 6 · Perf 6.5 · SEO 7.2 · AI 5.5 · Copy 8

The strongest proposition in the set, carried through every page rather than stated once. Patagonia reaches 8 on Proposition and Copy and sells the system around the jacket at the moment of adding it.

Where it did well. Proposition and merchandising together. Activism, repair and the Ironclad Guarantee run through navigation, hero and grid. Filters stack and show a live count against every option. The added-to-bag panel recommends a base layer, fleece and trousers, and the basket repeats them.

Room for improvement. The commit path. Apple Pay is offered in the basket and gone at checkout, and the fit sits in the product details rather than beside the size.

Watch-out. Patagonia does the hardest part of the category well and leaves the easiest undone. Moving one line of fit copy and keeping the express wallet through checkout would cost little and close most of the gap to first place.

67/ 100

3= Alpkit

Prop 6 · Exp 7.5 · Pers 7 · Chk 5 · Perf 7.6 · SEO 7.2 · AI 6.2 · Copy 7

A direct-to-consumer brand finishing above most of the household names in the category. Alpkit tops the Persuasion measure and carries some of the most distinctive product writing in the set.

Where it did well. Persuasion and copy. A 123-day Christmas returns window placed where a gift buyer hesitates, no pop-up, a category buying guide by hydrostatic head and weight, fit stated beside the size guide, and product tabs with a voice of their own.

Room for improvement. The recommendations. The Complete Your Kit panel shows an empty image placeholder, and the basket recommends a stove, a sleeping mat and a pump against a waterproof. Marketing consent is pre-ticked at checkout.

Watch-out. Alpkit sells a wider outdoor range than most of the field, and the basket is where that should pay off. Today it recommends the wrong part of the range, at the moment the customer is most likely to add a second item.

67/ 100

3= The North Face

Prop 6 · Exp 7.8 · Pers 6 · Chk 7 · Perf 6.5 · SEO 7.6 · AI 7.0 · Copy 5

Fit, model and size answered in one line, where the customer looks for them. The North Face leads the Checkout measure and is the most complete commit path among the top four.

Where it did well. The decision and the payment. “Relaxed fit” sits beside the model’s height and the size he wears, with a free size exchange stated where the doubt is. Benefit ratings, Shop the Look on the product page, and Apple Pay at the product page, basket and checkout.

Room for improvement. The machine-facing data. Seven of the eight Google merchant listings checked were priced in US dollars on the UK store. The basket recommends beanies and gloves in the low twenties against a £430 jacket.

Watch-out. Field data records a layout shift of 1.0 on mobile, the worst single reading in the set, under a respectable blended score. The page moving under the customer’s thumb as it loads is a conversion risk the headline number hides.

66/ 100

5. Rab

Prop 6 · Exp 8.6 · Pers 5 · Chk 4 · Perf 7.6 · SEO 8.0 · AI 8.0 · Copy 7

The best discovery in the set, with an unfinished commit path behind it. Rab tops Experience and shares the top AI Readiness score.

Where it did well. Discovery and data. Filters by outcome (most waterproof, most breathable, very light) alongside how-to guides, and clean, complete product data a machine can read.

Room for improvement. Everything after the add to bag. No recommendation at the drawer, the bag or the checkout. No express wallet at the product page, basket or checkout, a phone number required at the shipping step, and a prize-draw pop-up on the product page.

Watch-out. The customer who finds the right jacket through the best discovery in the set then meets one of the slowest routes to paying for it. The discovery work is the hard part, and Rab has already done it.

65/ 100

6. Gill

Prop 6 · Exp 8.0 · Pers 5 · Chk 5 · Perf 7.5 · SEO 7.6 · AI 8.0 · Copy 6

Built around sailing conditions, and let down by its own recommendations. Gill reaches 8 on Experience and shares the top AI Readiness score.

Where it did well. Category logic and fit. Categories by conditions (heavy weather, all weather, powerboating), a four-point fit scale under the size buttons, express payment in the basket and at checkout, and a full Agentic Browsing result.

Room for improvement. The Add to your system panel recommended a trouser available only in a 38-inch waist and a smock sold out in S, M and L. A £10-in-points pop-up fires on arrival, and marketing is pre-ticked at checkout.

Watch-out. The idea is right: a sailing jacket is bought as part of a system, and Gill says so. Recommending the system in sizes that have sold out turns a good idea into a dead end at the moment of highest intent.

64/ 100

7= Montane

Prop 7 · Exp 7.8 · Pers 5 · Chk 5 · Perf 7.8 · SEO 7.0 · AI 5.2 · Copy 5

Fast, specific and well merchandised in the basket, with consent handled badly. Montane scores highest on mobile Performance and runs the basket recommendations the category should copy.

Where it did well. Speed and product facts. The highest mobile Performance score in the set. Weight, fit and the model’s size under the size buttons. A basket that recommends waterproof trousers, base layers and an insulated jacket.

Room for improvement. Consent. The cookie banner offers no reject at the first layer, a 10% pop-up fires on arrival, and marketing is pre-ticked at checkout. A homepage line reads “challenges environments”.

Watch-out. The product data and the basket logic are the hard parts, and Montane has built both. The faults that hold it in the middle of the table are consent and copy, and every one of them is cheap to fix.

64/ 100

7= Henri-Lloyd

Prop 7 · Exp 7.3 · Pers 4 · Chk 5 · Perf 7.2 · SEO 7.2 · AI 7.2 · Copy 7

A sailing story told well, behind the most intrusive pop-up in the set. Henri-Lloyd finishes last on Persuasion.

Where it did well. Proposition and payment. The Originals Vintage archive and a sailing story dating to 1963. Shop Pay on the product page and express payment in the basket and at checkout, where the recommendations are a chino and T-shirt in the shopper’s own size.

Room for improvement. The arrival pop-up asks for a phone number and SMS consent with no discount attached. The basket recommends the same parka in another colour, and the community box at checkout is pre-ticked.

Watch-out. Checkout proves Henri-Lloyd can recommend the right things in the right size. The basket, one step earlier, recommends the jacket the customer has just added.

64/ 100

7= Dryrobe

Prop 6 · Exp 7.2 · Pers 5 · Chk 6 · Perf 6.7 · SEO 8.1 · AI 7.2 · Copy 6

A brand that owns its category, wrapped in sales mechanics. Dryrobe tops SEO Readiness and carries the richest shipping and returns data in the set.

Where it did well. Proposition and data. “The original changing robe” is a proposition a stranger understands at once. Sizing by height beside the size buttons, reviews shown inside checkout, and the only brand-written agent-instruction file in the set.

Room for improvement. A £10 pop-up fires before the cookie banner and returns as a tab after it is declined. The cart recommended car seat covers in options that had sold out.

Watch-out. Dryrobe’s customer is often buying a first robe as a considered purchase. Asking for an email before the cookie choice, and again after it has been declined, spends trust the category-defining proposition has earned.

63/ 100

10. Berghaus

Prop 6 · Exp 7.2 · Pers 5 · Chk 6 · Perf 6.2 · SEO 6.8 · AI 6.2 · Copy 7

A strong repair story behind a homepage that does not load cleanly. Berghaus reaches 7 on Copy.

Where it did well. The RepairHaus thread runs from the homepage to the product page (“You tear it up. We fix it.”), with good product-page voice and video in the listing.

Room for improvement. Nine of the 42 large homepage images failed to load. A 20% pop-up sits over the hero, the drawer recommends nothing, and the basket recommends competing jackets, including the same jacket in other colours. The homepage meta description is empty.

Watch-out. Blank image blocks on a homepage carrying a 40% sale are the first thing a Black Friday customer sees. They read as a shop that is not finished, in the week the brand most needs the customer’s confidence.

62/ 100

11= Jack Wolfskin

Prop 5 · Exp 7.8 · Pers 5 · Chk 6 · Perf 7.8 · SEO 7.6 · AI 5.0 · Copy 5

Seasonally consistent and clear, with recommendations that only sell more jackets. Jack Wolfskin shares the top Performance score.

Where it did well. Proposition and tooling. A winter hiking proposition carried through every homepage block, a size chart that opens in place and compares sizes side by side, a disclosure that some reviewers received a test product, and PayPal and Google Pay from the drawer onwards.

Room for improvement. Ten of the twelve basket recommendations were competing jackets. The fit filter tags 119 of 120 rain jackets as regular, and a member pop-up fires on the product page.

Watch-out. A customer who has chosen a jacket and is shown ten more is being asked to decide again. The recommendation slot is the cheapest place on the site to sell the second item, and Jack Wolfskin spends it on the first.

62/ 100

11= Fjallraven

Prop 7 · Exp 7.7 · Pers 5 · Chk 4 · Perf 6.5 · SEO 7.2 · AI 3.8 · Copy 7

Heritage identity and good fit information, then the hardest checkout in the top half. Fjallraven reaches 7 on Proposition and Copy.

Where it did well. Identity and fit. A distinct heritage, “Regular fit” under the product name, and a Fit Finder beside the size buttons.

Room for improvement. The commit path. Sign-in first at checkout, with guest as a secondary link, and no express wallet at any point. Filters took four to eight seconds to redraw after a click.

Watch-out. On the author’s walk the email step stalled several times before advancing to delivery. A customer who has already decided to buy should not have to decide again whether the site is working.

62/ 100

11= Mountain Equipment

Prop 6 · Exp 6.5 · Pers 5 · Chk 7 · Perf 6.8 · SEO 6.6 · AI 4.0 · Copy 7

The cleanest checkout gate in the set, behind a plain front end. Mountain Equipment shares the top Checkout score.

Where it did well. The commit path. Shop Pay and Google Pay in the drawer, the basket and at checkout, marketing left unticked, and a delivery date on every line of the bag. Feature-then-benefit specification copy.

Room for improvement. The front end. A homepage that reuses one line for two categories, a 10% pop-up over the product information, no recommendations anywhere, and a size guide the author called the weakest of the day.

Watch-out. Google could not parse Mountain Equipment’s product data. A brand with one of the best checkout gates in the set is one of the hardest of the twenty for a machine to read.

57/ 100

14. Columbia

Prop 5 · Exp 6.2 · Pers 5 · Chk 7 · Perf 6.2 · SEO 6.2 · AI 1.5 · Copy 6

Good ideas and mixed signals. Columbia shares the top Checkout score and holds the lowest AI Readiness score in the set.

Where it did well. Waterproofness levels (Rain, Downpour) used as a filter, fit stated beside the size buttons, and PayPal from the drawer onwards.

Room for improvement. Summer fishing and desert hiking imagery in a cold-weather campaign. A 10% pop-up on arrival, and beanies and caps recommended against a £200 jacket. The mobile page fails on layout stability.

Watch-out. The product page carries a Complete the Look outfit built around the jacket. The basket, one step later, recommends caps. The better recommendation already exists on the site and is not used where it would sell.

56/ 100

15= Helly Hansen

Prop 5 · Exp 7.2 · Pers 5 · Chk 4 · Perf 5.3 · SEO 8.0 · AI 4.5 · Copy 5

Strong review data, and a size chart that ends the journey. Helly Hansen sits joint second on SEO Readiness.

Where it did well. Fit consensus drawn from reviews sits beside the size buttons, with a size finder, and the store is structurally clean for search.

Room for improvement. The size chart is a separate page with no route back to the product. A 15% pop-up on arrival, no express wallet at any point, and no recommendations in the basket. Mobile responsiveness fails in the field.

Watch-out. A customer checking a size is close to buying. Sending them to a page they cannot easily leave, at that moment, is the most expensive place in the journey to lose them.

56/ 100

15= Regatta

Prop 4 · Exp 7.0 · Pers 5 · Chk 4 · Perf 6.5 · SEO 7.5 · AI 7.0 · Copy 5

A broad range, merchandised without a hand on it. Regatta scores well on AI Readiness and sits in the lower group on the customer-facing measures.

Where it did well. Filters that work, a free-delivery progress bar in the bag, and a broad range at accessible prices.

Room for improvement. The basket recommends women’s gloves, beanies and socks against a men’s jacket. Every size button reads “Select” to a screen reader. No express wallet, and the slowest field load in the set.

Watch-out. Two different recommendation styles on the product page (Style it with, and We recommend) ask the customer to work out which to trust. Recommendations aimed at the wrong customer are the clearest sign that nobody is merchandising the basket.

55/ 100

17= Craghoppers

Prop 5 · Exp 7.8 · Pers 5 · Chk 4 · Perf 5.5 · SEO 7.5 · AI 2.0 · Copy 5

A shop that sells before it explains. Craghoppers recorded the lowest mobile Performance lab score in the set.

Where it did well. A fit filter that works, and clear specification bullets on the product page including weight and back length.

Room for improvement. Add-to-bag buttons on homepage tiles, a membership pop-up on the listing and product pages, and a basket that offers a belt, a mug and socks against a £122.50 jacket. No express wallet at any point, and a membership box above guest checkout.

Watch-out. The product Q&A tab is empty and the delivery tab offers more options than a customer can compare at a glance. The store asks for commitment early and gives reassurance late.

55/ 100

17= Musto

Prop 5 · Exp 6.2 · Pers 5 · Chk 4 · Perf 6.5 · SEO 7.6 · AI 4.0 · Copy 5

Sailing authority, undercut by inconsistency. Musto’s discipline taxonomy is one of the clearest category structures in the set.

Where it did well. Categories by discipline (Offshore, Coastal, Inshore, Dinghy) and six decades of heritage carried on the homepage.

Room for improvement. The size chart is a separate page with no route back. The background and button styles change between the listing, the product page and the basket. A 15% pop-up on arrival, no recommendations in the basket, and no express wallet at the gate.

Watch-out. Google accepts a Musto product listing at a price of £0. A customer never sees it; an AI shopping agent comparing prices does.

53/ 100

19= Rohan

Prop 4 · Exp 6.3 · Pers 5 · Chk 5 · Perf 7.1 · SEO 6.1 · AI 4.0 · Copy 5

A clear travel heritage, sold almost entirely on discount. Rohan finishes last on SEO Readiness.

Where it did well. “For Every Journey” is a real proposition, the reviews summary is credible, and checkout offers PayPal, Apple Pay and Google Pay.

Room for improvement. Listing tiles carry New and 50% off together. The 15% pop-up fires again after it is closed. The basket recommends one T-shirt five times, and marketing consent is opt-out by default.

Watch-out. When most of the range is discounted, the discount stops being a reason to buy and becomes the price. A new product at half price on launch tells the customer something about the full price that the brand would not choose to say.

53/ 100

19= Paramo

Prop 4 · Exp 6.0 · Pers 5 · Chk 4 · Perf 6.5 · SEO 7.1 · AI 3.7 · Copy 6

Distinctive product, and a store that will not let the customer filter it. Paramo finishes last on Experience.

Where it did well. Performance-led product copy, an in-place fit and size guide, and the one basket recommendation in the set specific to how the product is cared for: the wash-in proofer it needs.

Room for improvement. Filter & Sort would not open in Chrome or Firefox on desktop or Chrome on mobile. The menu does not behave as a menu, and the homepage showed the words “No video available.” where its hero should be. No express wallet was seen.

Watch-out. Paramo’s fabric system is genuinely different from the rest of the field, and it needs explaining. A store that cannot be filtered, under a homepage that does not say what the brand is, leaves the explanation to a customer who has not yet been given a reason to stay.

The shape of the table

Help first. Ask later.

The instrument was built to expose a gap between reputation and execution, and in this category the famous names mostly hold their place. Arc’teryx and Patagonia lead on merit. The gap it found is a different one: between what these brands know about their customer and what their stores do with that knowledge.

Every brand in the set knows a waterproof is bought as part of a layering system. Four recommended the system in stock at the moment of adding the jacket. Every brand knows fit is the hardest question in buying outerwear online. Eight answer it beside the size buttons. And every brand knows a customer reading a product page is close to buying. Fifteen chose that moment, or the moment of arrival, to ask for an email address.

The finding of Edition Three

The highest-scoring sites help before they ask. They answer the fit question where it is asked, recommend the next layer rather than another jacket, and let the customer reach the product before requesting anything in return. None of that needs new technology. It needs the store to be merchandised with the same knowledge of the customer the brand already uses to design the jacket.

None of this predicts a brand’s sales. It measures something already true, on the day it was read, about the experience each brand chose to publish.

Read the full report

Edition Three, in full. Twenty-four pages.

Every measure explained, all twenty brands written up in rank order with where each did well, where it has room to improve and the single thing worth watching, plus the full methodology. Free, and there is no form in front of it.

PDF · 24 pages · 280 KB
DOWNLOAD THE PDF

Methodology

How the Index is built.

The Commercial Experience Index scores the digital experience a brand chose to publish, read entirely from the front end of a live site, out of 100. Every score traces to something observed on the store or in the objective data. Inference is labelled where it appears.

The eight measures

Proposition and Brand Clarity
Can a stranger tell what this is, whose it is, and why it, in three seconds.
Experience and Merchandising
Can the customer move without thinking, and is product merchandised with commercial logic.
Persuasion and Buyer Journey
Are prompts to buy present where the customer hesitates, and do they help or interrupt.
Checkout and Friction
The commit path, read as far as the front end honestly allows: express pay, guest gate, consent.
Performance
How quickly and cleanly the store loads and runs on a mobile device.
SEO Readiness
The structural conditions to be found, not rankings or traffic.
AI Readiness
Can a machine reach, read and correctly price the store from data the brand controls.
Copy and Voice
Do the words do a job, and does the brand sound like anyone.

Each measure is scored 1 to 10 from the live shopfront, and the eight combine into the total out of 100. Five measures are operator judgement; three read objective site behaviour. The balance between them, and how the total is composed, is the Barking Cat method and is not published.

The read standard

Every brand was walked twice against the same ten-point read standard: pop-ups and what they ask for, the homepage read to the footer, live navigation and filtering, badges on the listing, fit information and size guides on the product page, visual consistency from listing to basket, recommendations at three points after the add, and express payment at three points. Where the author’s first-hand walk and the standard read disagreed, the site was checked again; first-hand evidence stood where it held. Behaviour that appeared only under automated browsing was recorded and not scored.

The two ruling principles

Two rules prevent a specific inflation error, and both were applied uniformly to every brand. First, a supplier or technology badge is not the brand’s own identity work: a fabric lockup tells the customer what the jacket is made of, not who the brand is. Second, platform-default furniture never reaches the above-average band: the standard filters, quick-view and menu a platform ships out of the box are competent and no more. Credit goes to what the brand built on top.

The scale

Each measure is scored bracket-first. The centre of the scale is 5, the resting place of competent-but-unremarkable. A professionally run, unremarkable store scores in the 50s, not the 70s. The upper half is reserved for earned strength: real, visible work most competitors are not doing. The full range is live in both directions.

Scope note. This score reflects eight observable experience measures read from the live UK site on 17 November 2026: Proposition and Brand, Experience and Merchandising, Persuasion and Buyer Journey, Checkout and Friction, Performance, SEO Readiness, AI Readiness, and Copy and Voice. Performance draws on Google PageSpeed Insights mobile data for each homepage on the same day. Checkout is read up to the first personal-data field, with no account created and no order placed. Pop-ups were recorded with marketing cookies both accepted and declined; where the two differed, the shopper’s view with cookies accepted was used. The score does not include commercial diagnosis, sales, conversion or margin, organic search performance, paid media, stores, or AI output-presence (whether assistants recommend a brand when a real buyer asks them), which require data or live prompting beyond the URL and are covered in a full Barking Cat Review.

The Index is the front-end read a business can run on itself. The Barking Cat Review takes the whole store apart, shopfront and back office, establishes whether a fault is costing sales and roughly how much, finds the binding constraint, and puts commercial numbers against it.

Edition Two scored all twenty Premier League club shops on eight measures. Read Edition Two, or read Edition One, ten UK retail sites on seven measures.

Back to Thinking
The finding of Edition Three
“Help first. Ask later.”

The four highest-scoring sites never put a sign-up pop-up between the shopper and the product. Fifteen of the other sixteen did.

Barking Cat · Diagnosis before solutions