Bart Dunweg4 min read"It needs to be done from scratch." That is how a large share of our conversations with a company that already has a product begin. Sometimes it is true. More often it is half true: the product works, it just looks like 2017 and those two things get mixed up.
We have done enough redesigns ourselves to know what they cost. A completely new design for a product people use every day takes months, and during those months the old product stands still. Then the new one goes live, and the people who knew the old one by heart lose everything for a week. That is the bill. Sometimes it is worth it.
When it really has to start over
There are a few signals we cannot get around. Users fail a core task, not now and then but structurally; then something is broken in the setup and moving things around will not help. The foundation cannot take another feature without something else falling over. Or the company has become a different company, with a different brand and a different customer, and the product still tells the old story.
One test we always run: does the wish to start over live with the people who use the product, or only with the people who sell it? A sales team wants something new to show. A user wants it to work. If only the first is at play, a redesign is an expensive way to solve a marketing problem.
When you are better off building on
The product works and feels old. The numbers are sliding slowly, rather than dropping in one go. A new feature is coming that does not fit the current layout. In all three cases you build on. You take the screen where most people spend most of their time, get that right and learn from it for the next one. Every step goes live. The user notices it getting better, instead of suddenly being different.
Before we say that, we look first. We go through the product the way a user does, against real tasks, on the devices they actually have. What we find goes into a list, sorted by what it yields against what it costs. That list is usually longer than the company thought and shorter than it feared. And it answers the question we started with. If the top of it reads "confusing label" and "button in the wrong place" ten times over, you build on. If it reads "user cannot find the core function", you do not.
The middle, and what it looks like with us
We usually land in the middle. For a platform that has been running for years, we are currently building a new outer shell: the navigation, the overview, the parts that did not exist yet. The existing screens keep running inside it for now, in their old form, and move over to the new design one at a time. There is no day on which everything changes at once. The old product is maintained in the meantime, because people work in it.
Proving it got better
Whatever you choose, at the end comes the question: is it better now? For a product with hard numbers, like conversion or turnaround time, you measure those before and after. Many products do not have those numbers, or not yet.
Then we measure the experience. Before the redesign, a standard questionnaire to the users; after the redesign, the same one. For a client portal for healthcare practices the score came out almost twice as high, with the biggest gains on clarity and speed. That is the moment a discussion about taste stops. There is a number, and everyone in the room knows where it came from.
So if you walk in with "it needs to be done from scratch", the first question back is: what is not working right now? The answer to that decides the rest.
A thorough review of your product's user experience and interface that finds the friction and inconsistencies, and shows what to improve first.
