Four products, from B2B to B2C
Three years, four platforms, three audiences. I rebuilt an entire portal in under three months and went from guided execution to designing on my own. Those years taught me a trade no brief can hand you.
Edenred designs earmarked payment solutions: meal vouchers, gift cards, personal-services vouchers, mobility. Three audiences revolve around them.
The Clients are the companies that order, the Employees the people who spend, and the Merchants the shops that accept. I worked on the four platforms that serve them. I was the sole designer on the corporate site and part of a team on the other three.
Edenred.fr, the corporate site
Sole designer on the project.
The company’s showcase, and the main entry point for platform users. A site in constant evolution: page redesigns, photo swaps, banner and CTA tweaks, with an in-house team tooled up to make the simple changes itself.
There was no FAQ. People searched inside product pages or called the CRC directly, and a large share of its workload came from repetitive questions. And our catalogue is multi-product and multi-audience: the same question gets a different answer depending on whether you’re an employer or a beneficiary, on paper meal vouchers or on Kadéos.
I started with the classic FAQ: a list of questions and a search engine. Search was ruled out, technically out of reach. That constraint redefined the brief: I had to bring people to their answer another way.
That left the single list, which would have recreated the very problem it claimed to solve. And I couldn’t leave the sorting to the user: they don’t know our product names.
An “Edit” button on every step, to correct course without starting over.
"[…] In France many things are put in place but there is never any communication about them. Either you are a digger, you come across the information, or it goes unnoticed."A restaurant owner, November 2023 interview
Qualify before showing. You declare yourself in three or four steps (You are, Your solution, Your support, Your product) and then only see the questions that concern you. Two details hold it together: the steps arrive one at a time, never as a full form, and the selection trail stays at the top with an "Edit" on each step, so you can correct yourself without starting over.
A help center where navigation replaces search. After qualification, two levels are enough: themes on the left, questions on the right, the answer expanded in place, with no intermediate page. The system extends to each product world, each in its own color.
The help center was only part of the job: I also worked on the whole site’s layout, as reusable components declined from page to page, across a dozen product pages. The site has aged since.
The merchant area, for shop owners
Taking over a very old product: a full redesign, then steady iterations. Several designers over the years.
The portal for the shops and restaurants that accept Edenred vouchers in France. It’s the first merchant ground I worked on: a product nobody had really touched in years, and it showed. I moved it forward in stages.
It’s this work on the French portal that fed the next project: an international merchant area, designed with Edenred’s HQ to be rolled out across every country in the group. Turnkey for those without a portal, a replacement for those whose portal is limited.
In 2022, a complete redesign of the merchant partner area was greenlit, with less than three months ahead of us. A lot of pages and features to create, and a few key blockers to clear along the way. A high-pressure project, and one of the most absorbing of my Edenred France years.
The old portal’s mockups were unusable, all broken in an old design tool. So we started by rebuilding it in Figma, screen by screen, to work from a faithful copy of what was there. That base is what let us move fast afterwards: we knew exactly what we were replacing.
We shipped on time, and that’s what freed up the budget: the next phase could slow down and polish each journey and each screen before development. The portal has been iterating ever since. Three screens tell the story: the home page, invoices and the profile, each in three stages. What was there, my version, then the one the team deployed after I left.
"I sometimes feel that digital is more complicated than before. At the same time I understand, it wasn’t easy for them, it was new."A restaurant owner, November 2023 interview
The home page, in three stages
The starting portal: navigation scattered across tabs and shortcuts, with the merchant's activity buried among banners.
Same path for the two other everyday screens, invoices and the profile: left to right, the old screen, my version, then the one running today.
My Edenred, the app for employees
Designing new features on an existing base. Several designers over the years.
The mobile and web app where employees check their balance, manage their card and track their transactions. It’s the most used of the four, and the only one where the user comes back several times a week.
I designed new features on a base already in service, from mobile payment to blocking a card in an emergency.
The Plasticless journey dematerializes the card and its full map gives the real scope of the work.
From one card menu, you can trigger two neighboring actions with opposite consequences: suspend the card, or block it for good. One is reversible, the other isn’t. The risk I saw: someone rushed, card lost, panic reflex, triggering the irreversible while thinking they’re making a temporary move.
The button label can’t carry that difference on its own. What happens after the click must make obvious what you just triggered.
The stance taken: two journeys of opposite shape from the same menu.Suspend is a closed loop: the card is paused, and an explicit way back, "Reactivate my card", returns it to its original state. Nothing is lost. Block permanently is a one-way road that immediately carries the order for a new card: blocking is never a dead end, it leads into reissuing, then into what happens to the mobile payment already set up.
The mobile payment already set up is handled straight after, to bridge the wait for the new card.
Two journeys whose shape states their function: a loop for what can be fixed, a line for what has to be replaced. The user doesn’t confuse the two because their very architecture keeps them apart.
The client area, for companies
Designing new features on an existing base. Several designers over the years.
This is the most important of the four platforms: the one where companies order benefits and cards for their employees, and through which several billion euros pass. Everything we do downstream follows from those orders.
I worked on it as part of a team for several years, on pages already in service and much reworked since. Here is the shortest one to tell.
The MFA project took as much logic as screens: mapping every sign-in and activation case, e-mail then SMS, before drawing each step. The balance to strike: ask for the code often enough to protect, rarely enough not to wear out someone who just came to place an order.
The business before the screens
What I keep first from those three years is knowledge of the trade. I learned that an employer doesn’t always know the commercial name of their own solution, and that a question about the balance of a lost card gets a different answer depending on the format. You can’t make that kind of detail up: you have to stay long enough to know it.
I also keep a way of working: almost everything I designed was added to products already in service. I had to work with what was there, and with the work of the designers who came before me. It’s less comfortable than a blank page, and it’s what made me progress fastest.
And that ground opened up what came next: it’s because I’d understood merchants on the French portal that I could go on to design for every country in the group.
The outcome figures for this project belong to Edenred, so I don’t publish them here. What gets measured on the help centre is the share of repeat questions still reaching the support team.