A persona is mostly there to tell you what makes people close the door. Three parts of my work at Edenred: the personas and the workshop that opened the Deals feature, with the table that ruled one profile out of scope; the rituals that pace the team’s week; and the accessibility training I built for the designers and the product managers.
Deals started from a simple idea. A restaurant owner creates a promotion, it shows up in the app where employees look for lunch and he fills his tables in the quiet hours. Between him and that promotion sits a portal he only opens to download his invoices.
On paper, we already knew everything about them. What was missing was faces. I went back through the documentation we had and ran the benchmark. Out of that I built three profiles, and the workshop exercises around them. For each one I looked less at what they want than at what makes them close the door.
The ultra pragmatic
Solo restaurant owner, thirty-five years in the trade, four employees. Does his own accounts.
“If I don’t need it, why would I do it?”
What would move him
What shuts him down
The informed manager
Runs a restaurant, came from another industry. Her accountant often logs in for her.
“I don’t want to look like an idiot.”
What would move her
What shuts her down
The data-driven one
Several places, a structured network. Tracks her numbers every day.
“I already have tools that do this, what are you bringing me on top?”
What would move her
What shuts her down
Before designing anything I set three conditions a merchant had to meet to create a deal and I ticked them off profile by profile. Only two met them, and only partly. Dimitar met none of them: he left the workshop’s scope without leaving the research.
| The condition to meet | Persona 2Maria | Persona 3Elena |
|---|---|---|
| Sees Edenred as a business partner, not just a payment method | If supported | Almost |
| Sees the Edenred+ app as a customer acquisition channel | No | Yes |
| Is proactive on a platform she uses a few times a year | If guided | Yes |
| Ready to create deals? | Under conditions | Yes, but… |
acquired possible under conditions blocked
She takes the step if the interface tells her what to do and proves the gain. What was left to settle: what “worth it” actually means to her.
The most mature target to start with and the most demanding. A half-baked setup makes her leave for good.
“I accept Edenred because my customers use it” has to become “I use Edenred to generate business” The shift to obtain before adding any feature at all
That table produced the question for the session: where and how do we make Edenred’s commercial value visible before asking a merchant to create a deal?
The workshop brought two teams of three together over two half-days, with an outcome set from the start: three testable hypotheses and a handful of ideas precise enough to prototype.
Set out the problem, the scope and what we expect from the session. It’s the only moment I talk for long.
Everything we already knew, on the table before thinking. Everyone reads, nobody proposes anything yet.
The business framing, then the problems rewritten as How Might We, clustered by theme and voted on. We keep three.
The personas read out loud, then ideas in bulk. Crazy 8’s come last, by hand on folded paper.
The boards · made beforehand
Every exercise has its board: the instructions, the timing, examples to get going and an example of a bad answer next to the good one. Everything is ready before the session.
The production · two teams, two personas
One team on Maria, the other on Elena. The two piles of ideas have nothing in common: what reassures Maria irritates Elena.
The outcome · the ideas that won the vote
Regular meetings shape the team’s work. I ran or equipped them the same way: a board, instructions and a place set aside for each kind of remark.
Once a week someone puts their screen on the wall and the team looks at it. The board first asks what is being shared, where it stands (exploration, iteration, ready to ship) and what kind of feedback is wanted. Only then does the feedback arrive and it falls into three separate columns.
❓ The questions
What we don’t understand. We ask before we judge, which settles half the disagreements.
😍 What works
What has to stay. Without this column only the problems stick and we break what was fine.
💡 The leads
An idea, not an instruction. The designer stays in charge of the screen and decides afterwards.
When a subject starts, one session exists so that everyone leaves with the same version of the facts. This one opened the payment page redesign. Six blocks in this order. We don’t move on until the previous one is filled.
The goal
What we must have on the way out: the deliverables, agreement on the problem, the scope, the method, the schedule and the roles.
A · Why now
What makes the subject urgent. The people who carry it are the ones who fill this block.
B · The state of play
What we already know: the current journey annotated, the use cases, the data available and what we don’t know yet.
C · Scope & approach
What we touch, what we don’t touch and the various phases of work with their deliverables.
D · Roles
Who decides, who produces, who signs off.
E · Recap & next step
The questions left open and the date of the next check-in.
I took a two-day accessibility course and found it useful enough not to keep it to myself. I turned it into a one-hour session, open to anyone who wanted to come, and I built the whole deck. About fifty of us were in the room: designers, product managers, developers, every product role.
Condensing two days into one hour forces you to choose. I kept five parts: disabilities themselves, what has to be checked, how to install it in an organisation, then testing and auditing. Each board is a map you walk through on screen, with live demonstrations on real websites.
The session ends on tooling rather than on principles: the extensions to install in the browser and a questionnaire to place the organisation on the maturity scale. Everyone leaves able to check their own screen the next day.
Accessibility is useful to 100% of users. For 30%, it is necessary. For 10%, it is essential. The opening line of the training
What the team takes away
The Deals project, the rituals and the training all came from the same place: a moment when the team could no longer decide and something simple was missing to get the discussion going again. A sheet, an exercise board, a three-line table is often enough.
What I take from it is that a persona is mostly there to tell you what shuts people down, and that the table of three conditions still forces me to say who I am talking to before deciding what to build. The rest is a matter of preparation. I spend more time on it than on running the session, but when the board is ready we get into the subject in the first few minutes instead of spending a quarter of an hour explaining the exercise.