Image of the current homepage of bol.com

Why does bol.com wants to speak French?

Bol.com is an e-commerce site operating in the Netherlands and the Flemish parts of Belgium. In the past 20 years of existence, bol.com has grown to the number 1 webshop in that region. Bol.com is still growing and has the ambition to grow further. Going to the Walloon side of Belgium is a logical choice.

So Google Translate that shit, right?

Well, yes and no. Bol.com uses translation services from Microsoft Bing to translate the website into English. The problem is that machine translation has his problems. Often the translation is to literally. Bol.com’s promotions always have some kind of a joke in it that is local to the Netherlands or Belgium. That doesn't translate well. We also see that some content isn't translated because it isn't detected.
When bol.com makes a decision to expend to another region, they want to do it the right way. Therefore bol.com needs to invest in its technical environment to be able to handle other languages well. Something easier said than done.

Image of a error in machine translation. You see a warning sign with chinese writing on it. Below that the signs says in English 'Execution in process.' This sign is for work activities...

But how does bol.com wants to do it?

The app of bol.com is the first to support this kind of translation. This version will be released in the year 2020. Bol.com wants to do this step by step. The webshop will not be translated in the year mention before. I and some other UX designers saw some potential problems in the flow.

A big image of the not nice flows to bol.com app for customers of Walloon. The tricky part is that we have to redirect customers to the bol.com app, because that is translated to French. We expect that most customers will go to our website and see a Dutch site without any French translation. Not great.

0.1 flow

What we saw were two possibilities. When a Walloon sees a promotion on tv or an advertisement in a paper, the user can do two things. The user can open a browser and type www.bol.com and open the webpage. But sees a webpage in Dutch. Closes the webpage and have bad associations with that name.
The other flow is more positive. The user goes to the app store and download the bol.com app. When the user opens the app and gets the bol.com experience in their language.
We wanted to make sure that we handle the lousy flow as best as possible. We wanted to push the user to the app. Because there the user gets the best experience.

A image of a iteration of the flow to bol.com app. We show on our website a promotion that the app is in French. And the mobile flow from website to app is better than the desktop to app flow. It isn't great but better than the flow before.

0.2 flow

We realize that it is quite hard to push users from one medium to another. But after some discussions with the project, we came with an idea of how we can inform users that the website isn't useful yet and show the option of the app. With some kind of modal window with the necessary information to download the app.
It's still not the happy flow you wanted to facilitate, but this way, we can steer the user to a better experience. Because we had some doubts about this flow, we wanted to validate this flow with real users from Walloon. One problem none of us spoke French on a comfortable level. We contacted a research professional from that region to help us facilitate the research.

A image of the desktop and looking glass icons it represents the desk research we did.

Question part of the UX research

We wanted to validate as many assumptions we can do in the UX research. We wanted to see what users think when they see a pop-up on the entrance of a webpage. If a particular Walloon landings page was necessary? And how users would react when they do see automatically translated pages. To be extra sure that we need to facilitate our own translation in our pages.

Walloon landing page

My job was to design a Walloon landing page for the UX research. The primary purpose of this website is to convince Walloon to download the app. This page was also used to give the users some extra context what bol.com is, what the USP's are of the company, which features the app has, and some F.A.Q.'s to close the page. I did some competitor analysis, and you see that most pages have a clear call-to-action. Therefore, I have added a call-to-action three times on the page to provide an option to download the app when they want. An extra challenge is the French language itself. My assumption was that the words in the Dutch language were long. French, some words are even longer. Sort and scannable paragraphs were sometimes difficult. In collaboration with a visual designer, we made a final design.

Image of the wireframe of the landingspage for Walloon customers.
A big image of all the digital pages what we wanted to test. So we show two ways to inform users that they have to go to the app. In a bottom bar and in a typical pop-up. You also see the final design of the landingspage and the machine translated page.

Final digital design

When the date of the research came closer, we deliver our final designs. The UX designer couldn't make a decision about what kind of pop-up was most effective. The project group also didn't have a clear preference. So, we wanted to test them both. Next to that, you see my landing page and beside the machine-translated variant. You see a yellow bar where we communicate that this is a machine-translated page. We did this because we know from earlier research that users really like the head's up when the content isn't the best quality.

A image of the non digital part of the test. You see two types of flyers. We wanted to test the copy and promotion. To see what trigger most. And our bol.com box with Dutch word on it. We wanted to check how Walloon customers would react.

Final offline design

Bol.com is a subsidiary of Ahold Delhaize. Ahold Delhaize is a large company that is the owner of a lot of supermarket brands, among which the supermarket Delhaize. An idea of a project member was if we spread a flyer with some explanation of the bol.com app. In this flyer, we can also convert the user to the app with promotions like a discount on the first order or free shipping. We were curious if users liked these kinds of flyers.
Another thing we wanted to test is our shipping box with Dutch words on it. We wanted to see what the opinion was of the box. If the Dutch language had any impact on the perception of the brand.

Image of a room of me and some colleagues watching a Walloon customer from Liège interacting with the pop-up design.

Day 1: UX research in Liège

We had two days of research, one in Liège and one in Brussels. The first day was exciting because you noticed that Walloons are really different kinds of users, then users from the Netherlands. We had the idea that bol.com was really famous also in Walloon. But nobody knew us. They didn't know what kind of products we sell and were also not familiar with our propositions. We really need to build our brand perception in Walloon before Walloons trust us. Another big fail we had was that we used the word Walloon. This is a sensitive subject for people from Belgium. Belgium still is a land with lots of political difficulties based on the separation of Flemish and Walloon. Using Walloon is a big no, no.

Day 2: UX research in Brussels

Based on the feedback on day 1, we changed a lot of the designs. We stopped mentioning Walloon and visualized the assortment in the flyers, pop-up, and landing page. We immediately saw that this has some effect on the test. People were more positive and knew what kind of webshop bol.com is. Users didn't have any preference for how the pop-up should look. One likes the modal window, and the other liked the bottom bar. In this test, we saw that the landing page was more confusing than helpful. Because we say in the modal window that the website isn't in French and then we show a French landing page. So that wasn't that logical. It was two exciting days.

Image of a room of me and some colleagues watching a Walloon customer from Brussels interacting with the landingspage design.

So what now?

Bol.com is still in the middle of the process to support the French language in the app and web. All the insight we gathered we spread within the organization. These insights are used in decision making and gave some new ideas. I am inquisitive about what the end result will be.

What I learned from this project

I think this kind of project doesn't come often. Studying in a new culture and make a relevant webshop for that region. It was also a big reality check for me. In the Netherlands, bol.com has a brand recognition of 95%. We had the assumption that Walloon already kind of knew us. But we really have to start from the beginning of Walloon. That was something I didn't expect to be honest. It was a humbling process to see this. Something that I really keep in mind in my future work.

Curious for more content?

Project: Buyblock vision

About me