Give older listeners a second front door

Product teams are taught that good design is universal: build one product that works for everyone, and older or disabled customers will be served by the same clarity that helps everybody else. I believed that for most of my career, and I still believe it as a starting point. My position now is that for a consumer subscription with a large audience over seventy, universal design is necessary but not sufficient. Those listeners should be offered a second, deliberately reduced way into the same service, which they choose for themselves and can leave at any time, alongside a main app that meets accessibility standards in full.

I know how that sounds to designers. A separate mode for older people can look like a children's menu, and some of our own design team said so, forcefully. I want to set out why we built one anyway, what it cost us and where I think the critics are still right.

What the support line told us

About 19% of Lisandre's subscribers are sixty-five or older. They cancel less than any other group and listen to longer books, and in 2023 they generated 38% of all calls to our support team. I spend an hour every Monday listening to recorded calls, and for years the older callers' problems were small and specific: a book that had vanished from the downloads list, a password email that had gone to a junk folder, a sleep timer that would not switch off.

In September 2023 we released a redesigned player. It moved chapter navigation behind a swipe, replaced labelled buttons with icons and tucked the sleep timer into a menu. Testing had gone well. Our test panel, I realised later, contained eleven people over seventy out of three hundred. In the quarter after release, calls from subscribers over sixty-five rose by 60% and their monthly churn rose from 1.9% to 2.6%. Several callers told our agents that a grandchild had fixed the app for them and that they were afraid to touch it again.

The case against simplifying the main app

The obvious response was to simplify the main player for everyone. We prototyped a version of that, and it failed in the opposite direction. Younger and heavier listeners relied on the swipe navigation, the speed controls and the bookmarks, and removing them would have exchanged one group's frustration for another's. Our design director, Maëlle Courtois, argued that we should keep iterating until a single design served both. She was right that this is the better long-term goal. I judged that we could not reach it in time for the people who were already leaving.

We built Mode lecture, an opt-in mode with five controls, type at least twice the size of the main app's, one list of books in progress and a button that calls our support line directly. A family member can prepare a listening list from their own account. We introduced it in November 2024, offered it through support calls and through a printed letter to subscribers who had given a postal address, and switched nobody over automatically. To build it we set aside 22% of our product capacity for two quarters and postponed our free podcast tier from November 2024 to March 2025, which our finance director accepted with reluctance.

By mid-2025, 31% of subscribers over seventy were using Mode lecture. Calls from the over-sixty-fives had fallen 44% from their peak, and their monthly churn had returned to 1.8%. Not everyone was pleased. About one in eight people who tried the mode left it within a week, and some wrote to say they found the large type condescending. We reduced the default type size after those letters. I also accept a wider criticism: a separate mode can relieve the pressure on the main app to improve. We addressed that by giving our accessibility team a veto over releases of the main player, not only of the mode.

We also argued about whom to offer the mode to. The simplest rule would have been age, but we hold a birth year only for subscribers who chose to give one, and an age-based prompt would have insulted the many seventy-five-year-olds who use every feature of the main player. Instead the offer appears after behaviour that suggests difficulty, such as several failed searches for a book in the same week, or when a subscriber calls our support line. That choice made the mode slower to spread, and I still think it was the right one.

Accessibility law sets a floor, not a design

The European Accessibility Act, which applies to a range of digital products and services from June 2025, gave the main app a deadline. Our lawyers concluded that parts of our service fell within it, and we decided to treat the whole app as though it did. An external audit by Atelier Clairvoie in early 2024 found 412 issues against the accessibility standard we had adopted, from unlabelled controls that a screen reader announced only as buttons to progress bars that failed contrast requirements. By May 2025 the count was 37, all minor and scheduled.

The audit and Mode lecture solved different problems. The audit made the app usable with a screen reader and a keyboard and legible at high magnification, which matters enormously to our blind and partially sighted subscribers, many of whom were among our earliest customers. It did nothing for an eighty-year-old with good eyesight who simply did not want to learn a new gesture. Conformance with a standard tells a board that a product can be used. It does not tell the board whether the people paying for it can use it comfortably on an ordinary evening without ringing a grandchild.

I would make three suggestions to other product leaders and to the boards that oversee them. Measure older customers as a cohort in their own right, with their own panel members, support contacts and churn, rather than as a thin slice of an average. When those numbers move after a release, act within a quarter rather than waiting for a perfect universal design. And give the people who object to a separate door real authority over the main one, so that the second door never becomes an excuse for neglecting the first.