The global platform that flopped in Canada

WE’D EARNED THEIR TRUST on the air conditioner problem. We’d done the research, made the diagnosis, and redirected the budget toward things that worked. The client was happy. The relationship was solid.

But the website was broken in a completely different way.

It started as due diligence. We were reviewing the Midea and Toshiba Canada sites, both under our care by then, as part of getting familiar. We weren’t looking for an IA problem. We were just getting oriented. And within about twenty minutes, it was obvious why the digital ads weren’t converting.

The site nobody could use

The platform built in China had no filters.

Everything was commingled. Air conditioners, microwaves, humidifiers, dehumidifiers, toasters — all grouped together on category pages with no way to separate them. If you wanted to find a specific humidifier, you scrolled through everything in the small appliances section until you either found it or gave up. There were no subcategory pages. No brand pages. No product pages that actually felt like product pages. Just one long, undifferentiated list with no hierarchy and no way to narrow it down.

The client didn’t think this was unusual. They’d seen how Midea sold in Asia, through Alibaba-style marketplaces where the platform does the filtering for you. That model worked fine in markets where consumers were already familiar with the brand and the product types. In Canada, nobody knew who Midea was. They didn’t know what they were looking at. The site assumed the shopper already understood the brand landscape. Canadians didn’t.

Finding a humidifier meant scrolling past air conditioners and coffee makers and space heaters to get to the right section. Then finding the right size within that section. Three different models, all on the same page, no way to sort by capacity or price or features. It was an absolute nightmare. And the nightmare wasn’t theoretical. We could see it in the analytics. People were landing on product pages and bouncing immediately. The traffic had nowhere to go.

We showed the client what we were seeing. Not as a critique of their global platform. They hadn’t chosen it; they’d just been told to use it. But as a problem that was solvable.

The migration nobody expected

The fix was a full information architecture rebuild. We had to take the marketplace logic, category-less, filter-less, everything in one pile, and re-engineer it into what Canadian e-commerce shoppers expected: brand, then category, then subcategory, then product.

The WordPress rebuild went ahead. We built the category pages, the brand pages, the subcategory pages, and the product pages. We added the filters. The site stopped punishing people for visiting.

As we were finishing up the build, we joked with our local client that more often than not, HQ will make a decision to move platforms without a lot of local consulting to something like Adobe Experience Manager. This was not our first rodeo, and while we joked with the client that this was the next logical step, she did not believe it would happen. Or maybe she hoped it would not happen.

The client looked at us like we were the ones not seeing clearly. There was no way they were going to do it.

Two months later, we got the notification. In four months, they’d be migrating to AEM. They wanted something that could grow with the business instead of limping along on a DIY stack.

We didn’t argue. We helped them build the migration plan. We helped them scope the redirects. We made sure the new hierarchy carried over cleanly so they weren’t starting from scratch on the AEM side.

By the time the migration happened, the IA was already working. The category pages were organized. The product pages had the right structure. The filters were in place. The migration was a platform change, not a restructuring. The hard thinking had already been done.

What traffic could finally do

After that, everything else had somewhere to send traffic. The dishwasher launch. The laundry campaign that got parked pending inventory. The digital strategy had a foundation to work from.

The site stopped punishing people for visiting. But knowing people could find products was different from knowing whether the ads were driving them to stores. The third piece in this series will get into how we measured the visit, and why we still couldn’t measure the buy.