The brief

The client had launched an e-commerce site and invested heavily in it. Online sales were significantly disappointing, the investment wasn’t returning, and they wanted to know why.

What we found

The issues fell into three groups:

• Information architecture that hid the shop entirely.
• A user experience full of unexplained steps and broken-looking tools.
• Buyer journeys that didn’t fit how this audience actually purchases.

It wasn’t a customer problem, and it wasn’t a traffic problem. The site wasn’t converting the demand it already had.

What was working

• Strong organic traffic, much of it from people searching for specific products.
• Strong email engagement. The newsletters performed well.
• A content-to-product connection that set them apart. Published articles and abstracts linked research topics to related products.

The funnel didn’t match buyer intent

Only ~30% of visitors saw a product at any point in their journey. Of those who did and added it to their cart:

Adding to cart is a strong signal of intent, yet two out of three of those buyers never purchased.

• Almost all of the loss happened before checkout began. Once buyers started checkout, nearly all of them finished.
• That pointed to what buyers were (or weren’t) learning on the product and cart pages: that an account was required, who was allowed to purchase, and how institutions pay. The investigation started there.

Information architecture: the shop was hidden

There was no Shop link anywhere in the navigation.

• No menu mentioned products or shopping. You could only find products contextually, as related items deep inside research topics.
• Every competitor we reviewed led its navigation with Products, organized by product category.
• In usability testing, most people asked to find the shop couldn’t.
• The client’s reasoning was that buyers would read the research first and then purchase. That suits some buyers. Most arrived knowing exactly which product they needed.

Users were relying on search engines to navigate the site. The evidence:

• Organic search traffic landed on specific product names. People were using Google to reach products they couldn’t find on the site.
• “Acme Pharmaceutical + shop” was a common entry search: people knew the company sold products online, and couldn’t find where.
• The help page was one of the most heavily trafficked pages on the site.
• YOUR_CONFIRM: “help” was a common internal site search term. (Keep this line only if confirmed; delete it otherwise.)

User experience: friction at every step

The account requirement came as a surprise.

• Anyone could add to cart, but purchasing required an account. That wasn’t stated clearly anywhere, and nothing explained how to create one.
• The client hadn’t flagged the requirement as part of the funnel. It only surfaced during the discovery presentation, even though it shaped who could complete a purchase.

The product page template didn’t read as a product page.

• It was an out-of-the-box template made to fit. Key details sat far down the page and required heavy scrolling.
• Next to competitor product pages, the difference was obvious at a glance.

The shop filter looked broken.

• Selecting filters expanded the left rail to thousands of pixels tall, while the few matching results sat at the top of the page, out of view.
• The filter worked. It just looked like it didn’t.

Search only worked on exact scientific names.

• Partial or approximate product names returned nothing, and buyers rarely type the complete scientific name.

The help page used the wrong pattern.

• Each question sat in a card designed for promoting blog posts, so every answer opened on a separate page. Users had to click forward, check, go back, and try again.
• The page also mixed in other content types, which made it harder to tell what users were actually struggling with.

User journeys: the biggest buyers had no path

It wasn’t clear who could buy.

• These are pharmaceutical products sold to qualified buyers, not the general public. The site made it look like anyone could purchase.
• That also meant the add-to-cart numbers likely included unqualified visitors, so the demand the client believed they had may not have been accurate.

Institutional buyers had to leave the site to purchase.

• Most sales came through institutional procurement, tied to a purchase order and handled offline through email and the sales team.
• None of this was explained on the site. Buyers used it for discovery, then completed the purchase another way.
• Some site traffic was likely researchers planning purchases and sending links to their procurement teams. That activity couldn’t be tracked or attributed.

The channels weren’t connected.

• Email and the sales team drove most revenue, while e-commerce served colder audiences. The two didn’t connect.
• Existing customers had no reason or reward to buy through the site.

Why it happened

None of this came from a lack of effort. The team knew their products and their buyers, and they built a site that worked. What happened here happens on a lot of web projects: budget gets spread across many priorities, and user experience is often the piece that gets trimmed. A template looks like the efficient choice, and the team then spends the project making their business fit the template.

E-commerce is where that tradeoff costs the most. Every unexplained step between interest and purchase loses buyers, so the user experience has to be right, and it can’t be skipped.

A few common decisions added up:

• A complex purchasing process was adapted to an out-of-the-box template. Much of the functionality was made to work. But several parts of the journey needed custom builds the template couldn’t support.
• Information architecture was handled in-house to manage costs. The team brought real product knowledge to it. Without a user experience lens, though, the structure reflected how the organization sees its products rather than how buyers look for them.
• There was no dedicated UX review or testing before launch. Each piece was checked at a high level (“the filter works”). The issues that only appear when you walk through the site as a buyer went unnoticed.
• The business had outgrown its tracking and attribution setup. Connecting site, email and sales activity takes enterprise-grade analytics, and one person was carrying the site, email and tracking at once.

The optimization