Two words: French administration.

Have you ever encountered such an unassumingly innocent pairing of words that would leave you with all at the same time a sour, bitter and rotten taste in your mouth?

I spent close to six hours at the Police Prefecture queueing up for my bridging visa today. Can you imagine spending an entire workday of your life just to obtain a piece of paper?

As a person whose job is to use digital technology to improve experiences and processes, I thought there had to be a way to make that 6 hours even one hour less. Or less painful.

I could devise a theoretical case study of this part of the French administration process, but because of my lack of background understanding, any solutions proposed would be baseless. But I did learn some best practices and have come up with some steps to help you approach the task of solving wicked problems.

1) Choose the users who are the most vocal


When it comes to user research, besides sample size, you should also consider how powerful the feelings of your users may be.


I arrived at the prefecture at 10:15am. Joining the queue with me was a fairly vocal prefecture-goer, a tiny Moroccan beanie-wearing woman with a strong voice. It was her third time visiting, so she understood the ins and outs of the place, and also had some background knowledge of the paperwork procedures. She only needed to pick up her card which should literally take 2 minutes, but still had to spend 6 hours in line.

After 2.5 hours in the queue, an apparent injustice happened before everyone’s eyes where the security guard allowed a woman to skip the queue. Tiny Beanie was the first to ask her and the security guard why this was happening and to her disdain, which she afterwards communicated to the rest of us queuers.

We spent 3 hours queuing outside in the cold rain before they let Tiny Beanie and I inside. As I took my queueing ticket, I saw why the wait outside was so slow: the small waiting room was packed to the brim.

It was past lunch time, so I figured I’d go out to the nearest Franprix, grab a quick lunch and go back inside. When I returned, she welcomed me with a seat next to her and some fellow ladies. On my other side was literally a man sleeping across the next three chairs, snoring loudly; an accurate reflection of our collective mood.

It was here I realised that if anyone was going to try to improve the service experience by conducting primary research, you had to include Tiny Beanie in an interview because not only was she the most vocal and active member of the queuers, she was also the centre of a mini group of queuers sitting together discussing the situation. She had already spoken to other queuers and gave me a summary of what they were going through.  

Hence, when recruiting users for research, it’s worth including the users who tend to be more vocal and willing to share their feelings aloud. Not only will these users provide stronger feedback, they also prove to be more valuable than two or three users who may care less. These are the users who sacrificed a day of work to keep their papers updated. Therefore, not only is it about the quantity of users selected for user research, it also matters how strongly they feel about the subject.

Nielsen Norman Group’s study reinstates that the best results comes from testing with no more than 5 users in small tests, with 3-5 users in a category being the sweet spot. This is because once you add more users, you’ll start to see the same insights repeated again and again, making you learn less and less.

Ideal number of test users by Nielsen Norman Group


In summary, when approaching user research, there are two things to remember: Conduct user testing with small groups and be sure to include some of those more opinionated users.

2) On first impressions: the worst you can do is make them wait, so help manage their expectations

In real life settings, the second worst contributor to bad experiences after bad customer service is to make your customers wait. On a website or app, the loading time is the first experience we encounter before everything else.

We don’t consciously notice when a site loads fast, as we expect this phase to happen seamlessly. But when an app takes too long to load, it can seriously deteriorate your users’ first impressions. We are so habitualised to instant gratification, that on average 53% of us will abandon a site before it loads, according to Google (2016).

Slow load times and the bounce rate is proven to have a direct correlation. Once a user has to wait more than 3 seconds, the probability of bounce increases 32%.

In a physical queue, we may have a higher wait tolerance, but the tricky part is knowing when to leave a queue and cut our losses, since ‘bouncing’ out here isn’t as simple. That’s why in major amusement parks, you’ll often see a sign indicating how long the wait of the queue is estimated to be, because while queues are an inevitable part of going to Disneyland, we can only help control our customers’ perception of time. If I was told before joining the prefecture queue that the wait was 6 hours long, would I stay? Hell no. But not knowing this information made me feel as if I had no choice. That is a trapping feeling for anyone.

By tracking some data over a short period of time and some simple math, I assume that surely any facility or service centre should be able to readily provide wait time estimations, as a start.

3) And once we accept our fate, how might we be entertained?

Tell me if this is just a personal bias, but I’ve always classed boredom as one of the worst things one can feel. Maybe during my student days, I was lucky enough to even have the time to be bored, but now that I have work and responsibilities, you cannot just force me into boredom for one precious day of my life.

Airports have televisions, hairdressers and dentist clinics have magazines. Even shopping centres have play areas for children and resting areas for tired husbands. We live in a world that tries so hard to prevent our precious beings from feeling boredom.

So obviously a trip to an administrative office isn’t meant to be fun and maybe they have no wishes whatsoever to ease the process at all. But please, if you can’t fix the lagging areas of the system (even if it does save the government significantly in costs due to staggering inefficiency) at least make some efforts not to make it the place where your hopes and dreams die.

How I felt during a 6-hour queue for French admin


This anecdote tells a story of the ways our experiences can be greatly affected without even having touched on the main service. Been in a similar situation? Don’t hesitate to share in the comments below.