An adult toys retailer was losing most of its visitors before they ever saw the shop and nobody knew, because the culprit was the one element everyone clicks past without thinking. One month after we fixed it and three other things, conversion rate rose 173%.
| Client | Online adult toys retailer |
| Sector | E-commerce – adult / lifestyle |
| MARKET | EU |
| Engagement | CRO – behavioural research, A/B testing, iterative fixes |
| METHOD | Session recordings · heatmaps · A/B testing · analytics |
| TIMELINE | Results measured after one month |
Results after one month:
| Conversion rate | +173% |
| Sales | 8× |
| Homepage category CTR | +64% |
| Average pages per session | +34% |
| Product page views per session | +27% |
| Bounce rate | −31% |
The problem
The client had traffic. The client had product. The client was barely selling anything, and couldn’t see why.
This is the worst kind of conversion problem. Not a slow leak you can watch in a funnel, but a wall the analytics don’t show, because the people hitting it leave before they become a data point you’d think to examine. The dashboard said “high bounce, low conversion.” It didn’t say why, and no amount of staring at aggregate numbers was going to.
So we stopped staring at numbers and watched people instead. We ran session recordings – actual visits, replayed. And within a handful of them, the problem was obvious, sitting in plain sight in a place nobody thinks to look.
Why the obvious answer fails
Faced with “high traffic, low sales,” here’s what the site would normally get.
A homepage redesign. New hero, new colours, new copy. Expensive, slow, and aimed at a homepage most visitors were never actually seeing properly. You can rebuild the shop window all you like; it doesn’t help if the customer can’t get through the door.
More traffic. The reflex when sales are low: buy more visitors. But the visitors weren’t the problem. What happened to them was. More traffic through a broken entrance just means paying to fill a leakier bucket faster.
Pricing and promotions. Discount your way to conversions. This treats a symptom and trains customers to wait for the next sale and it does nothing for the people who were leaving before they ever saw a price.
Every one of these assumes the visitor is seeing the store and choosing not to buy. The recordings showed something none of them would have caught: most visitors never saw the store at all.
Solution 1: Fix the thing everyone clicks past
The culprit was the cookie banner.
It offered exactly one option: accept. No decline, no preferences, no close, no X. And here is what the recordings showed that no analytics report ever could: visitors didn’t accept it, and they didn’t dismiss it, because it gave them no obvious way to. They left it sitting there and it sat on top of the site, covering most of the shop. With the banner unresolved, only a small slice of the store was usable. Browsing was awkward and searching was worse.
So they did the rational thing: they left.
Two failures were stacked on top of each other, and both matter.
The design failure. A consent banner with only an “accept” button is a modal trap. It presents a decision and removes every way to make it except one, so users who don’t want to click the one button simply freeze, and a frozen user on a banner that occludes the page is a user who is now looking at a wall. The single most-ignored UI element on the internet was quietly hiding this entire business.
The knowledge failure. The recordings suggested people were also hesitating for a reason specific to this category: they assumed accepting cookies meant being followed around the internet by ads for an adult product, specifically. That fear is doing real work on an adult-toy store in a way it wouldn’t on a bookshop. Whether or not the tracking would actually have done that, the belief was enough to make people avoid the button and the banner gave them no other move.
We rebuilt the banner into an actual choice: accept, decline, manage preferences, close. The wall came down. For the first time, visitors could see and use the whole store and could resolve the privacy question without feeling cornered into surveillance they were, understandably, wary of.
This one change is most of why the numbers moved the way they did. Everything below helped. This is the one that unlocked the store.
Solution 2: Show people the products they’re actually looking for
With the store finally visible, the recordings surfaced a second, subtler pattern.
As visitors browsed, they kept navigating away from the categories featured on the homepage. The products on the front page weren’t the products people wanted. They’d land, not see what they were after, and go hunting and every extra step hunting is a step closer to leaving.
The homepage was merchandising by assumption. We replaced the featured categories with the ones the behaviour showed people were actually seeking.
Match the store to demand, don’t ask demand to search. A homepage’s job is to put the most-wanted things in front of the most people with the fewest clicks. When your data shows visitors consistently routing around your featured selection and they’re telling you exactly what to feature. We listened. Homepage category CTR rose 64%.
Solution 3: Put the important things where people actually look
We A/B tested the homepage layout, moving several sections higher up the page.
The detail of which sections won stays between us and the client. It’s specific to this category and this audience, and it’s the kind of thing that makes a store’s real advantage. The principle is general and worth having: position is persuasion. On the web, higher means seen, and seen is the precondition for everything else. Content that earns its place near the top gets attention that identical content, placed lower, simply never receives. We found what deserved to be higher and moved it there.
The result
A month is fast, and results this dramatic this quickly are rare. They happen for one specific reason: when a site has a critical fault – something actively blocking sales rather than merely dampening them – removing it produces a jump that a healthy, well-built store could never show, because a healthy store has no such block to remove. Fast, enormous gains are not a sign of routine brilliance. They’re a sign the starting point was badly broken. We were fortunate to find a problem this severe and this fixable. Most engagements are grind, not fireworks.
And “sales 8×” needs an asterisk we’ll add ourselves. When the cookie banner was hiding the store, sales were close to nothing so the baseline was artificially floored. Multiplying a very small number gives a spectacular multiple that sounds bigger than the underlying reality. We didn’t conjure eight times the demand out of thin air. We removed a wall that was suppressing demand that already existed. The honest way to read this case is not “we made a good store great.” It’s “we found a store that couldn’t sell, and let it.”
That’s why conversion rate – up 173% – is the number to trust here. It’s a rate, not a raw count, so it’s far harder to flatter with a low baseline, and it measures the thing that actually changed: of the people who arrived, dramatically more of them could now become customers.
If you’re in the same position
Watch real sessions before you redesign anything. Aggregate analytics tell you that people leave. They rarely tell you why, and the why is often something no dashboard would ever surface – a banner, a broken button, a trap nobody on the team ever hits because they all click “accept” on reflex. A few recordings can be worth more than a quarter of A/B tests.
Your consent banner is part of your store, not a legal afterthought. If it offers only one option, it isn’t asking for consent, it’s blocking the page until surrender. And a banner that covers your shop is your most-viewed and least-examined page. Give people a real choice, including the choice to say no. The ones who felt cornered were the ones leaving.
Merchandise to the demand you can measure, not the demand you assume. When visitors consistently navigate away from what you’ve featured, don’t take it personally. Take the hint. They’re telling you what to put on the homepage.
The most expensive problems are the ones nobody’s looking at. This business was losing most of its sales to an element every single person on the team clicked past without a thought, every day. It wasn’t hard to fix. It was hard to notice and noticing is the job.