A joinery business in Horsham spent the best part of a fortnight comparing Shopify against WooCommerce, reading forum threads about transaction fees and app costs, before finally settling on a platform. Eighteen months later the shop was still turning over less than the owner's old eBay listings, not because the platform was wrong, but because nobody had touched the product pages, the checkout still asked for a phone number nobody wanted to give, and there wasn't a single customer review anywhere on the site. The platform decision took two weeks. The decisions that actually mattered took eighteen months to get wrong.
Why the platform you picked was the easy decision
Choosing between Shopify, WooCommerce, Wix or Squarespace is typically a one-day decision once you sit down and actually compare them. There's a shortlist, a few YouTube walkthroughs, a free trial, and a winner. What happens in the following twelve months, the product page tweaks, the checkout fixes, the slow build of trust with strangers who've never heard of your business, is what decides whether the shop makes any money at all.
Most builder comparison articles, and there are dozens published every year ranking the latest platforms, focus on templates, pricing tiers and app marketplaces. Understandably so; that's the decision people are searching for help with. What they rarely cover is what happens after the payment gateway is switched on and the first stranger lands on a product page with a card in hand and thirty seconds of patience.
UK ecommerce conversion rates typically sit somewhere between 1.5% and 3%. That means even a well built shop, on the right platform, with decent products, loses 97 or more out of every 100 visitors without a sale. The platform didn't cause that. Product pages that don't answer basic questions, checkout friction that makes people give up at the last screen, and a lack of anything that builds trust in an unfamiliar seller usually did.
The gap between a shop that survives and one that grows is rarely the platform logo sitting in the footer. It's product pages, checkout friction and payment trust, the unglamorous, ongoing work that never makes it into a builder comparison table. Everything covered from here applies whether the site runs on an off the shelf builder or a bespoke build, the kind Samson Web Design has built for Sussex and UK small businesses since 2006. The principles don't change with the software.
Put a number on it and the stakes become clearer. A typical UK small business running Google Ads for a competitive product term might pay somewhere between £1.20 and £2.50 per click. Send 500 clicks a month to a product page converting at 1.5% and that's roughly seven or eight sales for £750 to £1,250 in ad spend, a cost per sale north of £100 before the product itself is even paid for. Nudge that same page to a 3% conversion rate, entirely achievable with the fixes covered below, and the same ad spend produces fifteen sales instead of seven, roughly halving the cost of acquiring a customer without touching the media budget at all. Nobody talks about this in a builder comparison table, because it has nothing to do with which platform was chosen.
Three free tools are worth installing on day one, regardless of platform. Google Analytics 4 shows exactly where in the funnel visitors drop off, product page, cart, or checkout. Microsoft Clarity, free and simple to set up, records anonymised sessions and heatmaps so you can watch someone scroll, hesitate over a button, or rage-click a delivery link that didn't work. Google Search Console shows which search terms actually bring people to specific product pages, which matters more for deciding what to fix first than any amount of guesswork. None of the three cost anything, and together they matter more to conversion than most paid app marketplace add-ons ever will.
Product pages that do the selling for you
For most shoppers the product page is the real landing page, not the homepage. They arrive from a Google search, a Facebook ad or a link shared in a group chat and land straight on that one page, often having never seen the homepage at all. If that page doesn't do the job of a shop assistant, the visitor leaves and the homepage never gets a chance to make a first impression.
Structure matters more than most sellers think. A title that uses the words customers actually search for, not an internal product code or a cute brand name. A clear price, not a range that requires a click to resolve. Honest stock status, whether that's "12 in stock" or "back in stock 14th". And an Add to Basket button that's visible without scrolling, sitting where the eye naturally lands after the price.
Research into small business web design backs this up directly: calls to action need to sit prominently on product and checkout pages, not buried under paragraphs of copy or pushed below a fold. It sounds obvious written down. It's astonishing how often it's ignored in practice, usually because the description was written first and the buy button was added as an afterthought underneath it.
A Sussex homeware seller Samson worked with had exactly this problem. Their Add to Basket button sat below a long, well written description, itself below the fold on most screens. Moving the button to sit directly under the price and product image, with the fuller description pushed further down the page where interested buyers could still find it, produced a noticeable rise in add to cart clicks within a matter of weeks. Nothing else on the page changed. Just where the button lived.
Put a number on that rise: add-to-basket clicks moved from roughly 2.1% of product page visitors to 2.8%, a jump of about a third, measured over six weeks with nothing else on the page touched. That's the kind of change that, multiplied across a whole catalogue, is worth more than months of platform research ever could be.
A fifteen-minute audit, done on your own site right now, catches most of this without needing any tools at all. Open the site on a phone, not a desktop, because that's how most of your traffic arrives. Find your five best-selling products and, for each one, time how long it takes to answer three questions just by looking: what does this cost including delivery, is it in stock, and how do I buy it. If any answer takes more than a couple of seconds of scrolling or a click through to another page, that's the fix to make first, ahead of anything else on this list.
The edge case that catches almost everyone out is exactly that gap between desktop and mobile. A button that sits neatly under the price on a wide desktop monitor can end up three or four scrolls down on a phone screen, once a long title, a size selector, a stock message and a delivery estimate have all stacked up above it. A trade supplies site Samson reviewed had a perfectly placed button on desktop and a button sitting below two and a half phone-screens of scrolling on mobile, where the large majority of its traffic actually arrived. Google's own Mobile-Friendly Test and PageSpeed Insights, both free, will flag layout and speed issues, but neither will tell you where a button actually sits on a real phone screen. Only checking on an actual phone catches that.
Delivery estimate, size or fit guidance and stock levels should live on the page itself, right there next to the button, not force a click through to a separate shipping policy page that most visitors never bother opening. Every extra click is a chance for someone to lose interest, get distracted by a notification, or simply decide it's too much effort to find out if a jumper runs small.
The classic trade website mistake is a page that says little more than "contact us for details" where a price and a buy button should be. It feels safer to the business owner, who's used to quoting jobs over the phone. But that single choice quietly turns browsers back into phone callers at best, and far more often, into lost sales to a competitor two search results down who simply listed a price.
Photography that replaces touch and try-on
Online shoppers cannot pick a product up, turn it over, feel the weight of it or check how the fabric moves. Photography has to do the job that a shop assistant or a fitting room mirror would normally do in person, and most sellers ask far too little of it.
There's a minimum standard worth holding every product to. Several angles, not just the one flattering shot. A working zoom on the main image, so a buyer can check stitching, texture or finish up close. And at least one lifestyle or in-use shot that shows scale and context, a chopping board on a real kitchen counter rather than floating on white, a jacket on a person rather than a hanger.
None of this needs a studio. A £30 to £50 lightbox and a modern smartphone used in good daylight, ideally near a window rather than under a ceiling bulb, will comfortably outperform a single dim photo taken on a stockroom shelf with the flash on. The equipment cost is trivial next to what it does for conversion.
A handful of free or near-free tools cover most of what a small seller needs beyond the camera itself. Photoroom, a phone app, strips a busy background out and drops in plain white or a chosen colour in seconds, useful for sellers who can't easily set up a lightbox for every item. Canva turns a plain product shot into a lifestyle-style graphic with dimensions or a size guide overlaid, handy for clothing and furniture. Before anything gets uploaded, running it through a free compressor like TinyPNG or Squoosh matters more than most sellers realise: a single uncompressed photo straight off a modern phone can sit at 4MB or more, and a product page with six of those loads painfully slowly on mobile data, where a large share of shoppers still are. Google's own research puts the cost plainly: 53% of mobile visitors abandon a page that takes longer than three seconds to load, and every extra photo left uncompressed adds to that delay.
A Sussex furniture seller Samson worked with had exactly this problem without realising it. Product pages carried eight to ten full-resolution photos each, some over 5MB, pulling mobile load times past six seconds on a typical 4G connection. Compressing the same images to under 300KB each, with no visible loss in quality on a phone screen, cut load time to under two seconds and coincided with a measurable drop in the page's mobile bounce rate over the following month.
Short video helps too, and it doesn't need to be polished. A fifteen second clip filmed on a phone, showing a tool being used, fabric moving as someone walks, or steam coming off food, noticeably helps conversion for anything people would normally need to see in motion to understand. It answers a question stills simply can't.
What matters just as much as any single photo is consistency across the whole catalogue. Same background, same angle, same lighting style on every product. That consistency reads as professional and trustworthy to a visitor even when the individual shots are simple rather than glossy, whereas a mix of styles, some on white, some on a wooden table, one clearly taken in a different decade, reads as amateur no matter how good any one photo is on its own.
Descriptions that answer the question the customer hasn't asked out loud
Every product description should answer three silent questions before the customer scrolls away: what is this, who is it for, and why should I trust it. Most descriptions on small UK ecommerce sites answer none of them properly, because they were written in five minutes to fill a blank box rather than to persuade someone with a card in hand.
Copying manufacturer text word for word is the most common shortcut, and it causes two separate problems. Search engines see the same paragraph on dozens of other sites and rank it poorly because there is nothing unique to reward. Customers, meanwhile, skim past it because it reads like every other listing they've already scrolled through that afternoon. Neither problem gets fixed by tidying the formatting; the words themselves need to be yours.
Two competitors selling the identical camping stove, both running the exact same manufacturer paragraph pasted straight from the supplier's spec sheet, will typically both sit further down the search results than a third seller who bothered to rewrite even a hundred words of it in their own voice. Google can tell the difference between original description and duplicate copy, and it rewards the former even when the product itself is identical down to the model number.
The fix is to write to the specific worry a UK buyer actually has about that exact product, not a generic worry a copywriter invented. That might mean sizing translated into UK measurements rather than EU or US ones, material and washing instructions spelled out in plain English, compatibility with kit the customer already owns, or a realistic delivery timing statement rather than a vague "usually ships in 3-5 days". These are the details that turn a browser into a buyer, because they remove the need to go and check somewhere else.
A workable structure that consistently performs well is short and repeatable. Start with a benefit led opening sentence that says what the product does for the customer, not what it is made of. Follow with a bullet list of the hard specifics: dimensions, materials, compatible fittings, whatever matters for that category. Then close with a short paragraph that tackles the one question you already know generates the most pre-sale emails or phone calls, because you've been asked it a dozen times before.
This last point matters more for trades and services than almost any other business type selling online. A plumber selling parts online should write the description the way they'd explain the part over the phone to a customer who doesn't know a compression fitting from a push-fit, not copy the spec sheet straight from a supplier catalogue. Jargon that sounds professional in a trade counter conversation reads as a wall of noise on a screen, and it's the fastest way to lose someone who was ready to buy thirty seconds earlier.
A before and after makes the difference concrete. Before: "15mm x 1/2 inch compression tee, brass, WRAS approved, dual thread configuration." Every word accurate, and almost useless to a homeowner trying to work out if it fits their existing pipework. After: "Joins a 15mm copper pipe (the standard size for most UK domestic plumbing) to a 1/2 inch threaded fitting, so you can add a tap, valve or radiator branch without cutting into new pipe. WRAS approved, meaning it's certified safe for drinking water systems." Same part, same accuracy, and one version a non-plumber can actually act on without ringing to ask.
Free tools cover most of what's needed to write these well. AnswerThePublic and Google's own "People also ask" boxes show the exact phrases real customers type when they're unsure about a product, which is where the closing paragraph's question usually comes from. The Hemingway App flags sentences that have got too long or too dense to skim on a phone screen. And on WooCommerce specifically, a free plugin like Yoast SEO or RankMath scores a description for basic search readability without needing any technical knowledge to interpret the result.
Reviews and social proof that close the trust gap
Research cited by Klaviyo UK found that 74% of shoppers say product reviews are a key way they learn about products they've never bought before, particularly from brands they don't already know. That single statistic explains why an independent UK seller without national brand recognition needs reviews more urgently than a household name does, not less.
For a small trades or startup site, reviews do more persuasive work than any amount of polished design, because they substitute for a trusted friend's recommendation. A visitor has no personal history with your business, so they borrow someone else's. Slick photography and a clean layout build a baseline of credibility, but reviews are what actually close the gap between "this looks professional" and "I'll buy from these people".
Gathering them doesn't need to be expensive or complicated. An automated post-purchase email asking for a quick review a week or two after delivery, timed to land once the customer has actually used the product, is the single most reliable method. A Google Reviews widget pulled through to the site, or a free Trustpilot business account, both work well and cost nothing beyond the time to set them up.
The practical steps are worth spelling out because most sellers skip the timing and wonder why response rates are low. Send the request seven to fourteen days after delivery, not on the day it arrives, so the product has actually been used rather than just unboxed. Keep the email to two short lines and one clear button, not a long thank-you letter with the review request buried at the bottom. And ask a specific question rather than a generic one, "how did the jacket handle the rain this week?" gets a longer, more useful answer than "please leave us a review". Paid platforms like Trustpilot's automated tiers run from roughly £199 to £399 a month for a small business, but free alternatives, a scheduled email through whatever platform the store already runs on, plus a Google Business Profile request link, do most of the same job at no cost.
A negative review will arrive eventually, and how it's handled matters as much as the review itself. Responding publicly, briefly, and without getting defensive reads well to every future visitor who sees it, not just the reviewer. Something close to "Sorry this didn't work out, we've refunded in full and would welcome the chance to put it right, please email us directly" does more for trust than a five-star average with no replies at all, because it shows a real business behind the shop front rather than silence.
User-generated content, the ordinary photos customers post of a product in actual use rather than staged studio shots, tends to persuade more than star ratings on their own. A photo of a garden ornament sitting in someone's actual flowerbed, wonky fence and all, does more to convince a hesitant buyer than a five-star average ever will, because it shows the product surviving contact with real life.
Fake or incentivised reviews are worth avoiding entirely, and not only on ethical grounds. Following Competition and Markets Authority enforcement action, and with new powers under the Digital Markets, Competition and Consumers Act 2024 allowing the CMA to fine businesses directly rather than going through the courts, manipulating reviews is a genuine legal risk for UK businesses, not a grey area. Slow, genuine accumulation is both the safer route and, in practice, the more convincing one, because real reviews read differently to fake ones and customers have got better at spotting the difference.
The result of doing this patiently is straightforward and provable: a small trades site with twenty honest, specific reviews will consistently outsell a slicker competitor with zero social proof, regardless of who spent more on design.
Checkout: where UK shoppers abandon and why
Industry data consistently shows that around seven in ten online carts are abandoned before purchase, and a large share of that isn't simple price shopping, it's friction that appears the moment someone tries to actually hand over money. This is the point in the journey where a well-built product page and a good review section can still lose the sale, because the checkout itself gets in the way.
Research from the Baymard Institute, which studies checkout usability across thousands of live ecommerce sites, breaks the reasons down in a way worth knowing before assuming price is always the culprit. Extra costs (delivery, taxes, fees) appearing too late in the process accounts for roughly half of all reported abandonment. Being forced to create an account accounts for around a quarter. A checkout process that's simply too long or too complicated, and a lack of trusted payment options, make up most of the rest. None of those four are about the product being too expensive; all four are fixable without changing a single price on the site.
Forced account creation is one of the most common reasons UK shoppers give up entirely. Being told you must register, choose a password, and confirm an email address before you're allowed to buy a £15 item feels like an unreasonable ask, and plenty of customers simply close the tab rather than comply. Guest checkout should always be offered as the default, easiest path, with account creation presented afterwards as an optional convenience for next time, never a gate before purchase.
Keeping the number of steps low matters just as much as removing the account wall. Whether the flow is a single page or a short multi-step sequence, a visible progress indicator, something as simple as "Step 2 of 3: Delivery", reduces the feeling of an open-ended process that could demand yet more information at any moment. Customers tolerate a few steps far better when they can see the end of them.
Mobile checkout has its own edge cases that catch small sellers out more than desktop ever does, simply because most owners test their own site on a laptop. Address autofill, using a free service like Google's Places API or a UK-specific tool like Loqate, lets a customer type the first line of an address and select the rest, rather than typing a full postcode, town and county on a phone keyboard one field at a time. Card number and postcode fields should trigger the numeric keypad automatically rather than the full letter keyboard, a small setting most platforms support but few store owners ever check. And the classic mistake, a phone number field that insists on a UK landline format and rejects a mobile number starting 07, quietly blocks a meaningful share of checkouts on a device most people are actually using to shop.
Cost transparency is where a lot of otherwise decent checkouts fall down. Showing the full cost, including delivery, as early as possible in the journey, ideally on the product or cart page, matters far more than revealing it as a surprise on the final screen. An unexpected delivery charge appearing right before the payment button is one of the single biggest causes of abandonment, because it feels like the price changed at the last second even when technically it didn't.
The payment gateway sitting underneath the checkout is worth a moment's thought too, separate from how the checkout looks. Stripe and PayPal both charge UK small businesses roughly 1.4% to 2.9% plus 20p to 30p per transaction depending on card type and volume, with no monthly fee on standard tiers, which makes either a reasonable default for a small seller rather than something to agonise over. The bigger decision is which of them, or both, to offer side by side, because familiarity with the payment brand itself is part of what makes a stranger comfortable entering card details on an unfamiliar site.
Advice drawn from first-hand experience of building ecommerce sites is refreshingly blunt on this point: resist over-engineering the cart and payment flow. There is a real temptation to add extras, saved wishlists, complex upsell modules, multiple delivery options with their own sub-menus, in an attempt to compete feature-for-feature with the likes of Amazon. For a small UK seller this is almost always wasted effort; the better move is to keep the flow as simple and predictable as possible rather than chase complexity nobody asked for.
The clearest illustration of this came from a Worthing-based client whose checkout included a mandatory account creation step. Analytics showed a sharp, consistent drop-off exactly at that screen, week after week, regardless of what else changed on the site, with roughly six in ten visitors who reached that page leaving without completing it. Once that single mandatory step was removed and replaced with a guest checkout option, the drop-off at that point in the funnel fell to within a few percentage points of every other step in the process, with no other change to the page. It was a reminder that checkout problems are rarely mysterious; they are usually one obvious obstacle that nobody had got round to removing.
Payment trust: the last hurdle before someone taps pay
Everything up to this point gets a stranger to the payment screen willing to buy. What happens on that screen, and around it, decides whether they actually finish. This is a distinct problem from checkout friction, which is about how many steps and fields stand in the way. Payment trust is about whether, at the exact moment a card number is being typed in, the site looks like somewhere that number is safe to hand over.
Trust badges are the most visible version of this, and also the most misunderstood. Research on checkout behaviour, including studies from the Baymard Institute and Nielsen Norman Group, consistently finds that generic security badges, a padlock icon, a vague "100% secure" graphic with no recognisable name attached, do almost nothing on their own. What does move the needle is a logo a shopper already recognises and already trusts from using it elsewhere: the actual PayPal, Visa, Mastercard or Apple Pay marks, shown as accepted payment options rather than as decoration. A shopper doesn't need to trust your business yet to trust that Visa's fraud protection exists; borrowing that established trust is far more effective than any badge invented for the purpose.
Offering more than one payment method matters for a similar reason, and it's measurable rather than just a feeling. A shopper who prefers to pay with PayPal because they don't want to type card details into a small business site they've never heard of will simply leave if card is the only option on offer. Adding Apple Pay and Google Pay removes typing altogether on mobile, which is where a large share of UK ecommerce traffic now sits, and both show up as an instantly recognisable payment icon that needs no explanation.
Buy-now-pay-later options, Klarna and Clearpay being the two most recognised in the UK, deserve a mention specifically because of how differently they land depending on the product. For anything over roughly £50, offering the option to split payment into three or four instalments, at no extra cost to the customer, noticeably raises average order value for categories like furniture, clothing and homeware, because it lowers the psychological barrier of a single large payment. A Sussex homeware seller adding Klarna as an option alongside card and PayPal saw average order value rise measurably within the first two months, with no change to pricing, simply because customers who'd have hesitated at £180 in one go were comfortable with three payments of £60. It's worth noting the FCA brought BNPL providers under regulated consumer credit rules from 2026, so it's sensible to check a provider's current compliance position before adding one, but as a trust and conversion tool it remains one of the highest-impact single additions available to a small seller.
Company information sitting quietly in the footer does more legal and psychological work than most sellers realise. Under the Consumer Contracts (Information, Cancellation and Additional Charges) Regulations 2013, a business selling online in the UK is legally required to make its business name, registered address, and contact details available to the customer before a purchase is completed, not buried in a terms page nobody reads. Beyond the legal requirement, a visible company number, a real phone number, and a physical town or city, even just "based in Worthing, West Sussex" rather than no location at all, reassures a stranger that there's an actual business behind the shop front, not a website that could vanish the day after a payment goes through.
A clearly stated returns policy, visible before checkout rather than discovered afterwards, closes a related worry. UK consumers buying online already have a legal right to a 14-day cooling-off period under the same 2013 regulations, during which most goods can be returned for a full refund. Stating that plainly on the product page or in the checkout, rather than leaving customers to assume it or go searching, removes one more reason to hesitate at the final screen. Sellers who bury this information, or worse, display a policy that tries to offer less than the law requires, tend to see it surface as a source of chargebacks and disputes later rather than lost sales up front, which is a more expensive way to learn the same lesson.
HTTPS and a valid SSL certificate are close to universal now, and every mainstream platform provisions one automatically, but it's worth checking rather than assuming, particularly on a custom domain freshly connected to a builder. A browser warning reading "Not Secure" next to the address bar at the exact moment a customer is about to enter card details ends the sale instantly and irreversibly, regardless of how good everything else on the page is. It costs nothing to check and takes thirty seconds; the cost of not checking is every sale from every visitor who sees that warning.
