Vulnerability detection: turning questionnaires into genuine care

September 4, 2026

by Ben Leonard

CEO & Co-Founder at Life Moments

Most banks and insurers have come a long way on vulnerability. Nearly all now run a dedicated support function, train their frontline teams, and do careful work for customers who tell them they are struggling. That work matters and deserves more credit than it gets. Almost all of it, though, rests on one assumption: that the customer will raise their hand.

The FCA’s Financial Lives survey shows how much that leaves out. In May 2024, 26.4 million adults, over a third of the country, showed at least one characteristic of vulnerability. Yet only around four in ten people in vulnerable circumstances have told their provider. The lack of visibility on vulnerability is the detection gap, and it is where we encounter real risk to good outcomes.

Why asking does not work

The gap is not a failure of effort. It is built into the method. Ask people to declare a difficulty and most will not. Among those who had not told their provider, the FCA found that 37% felt embarrassed, a quarter did not want to be treated differently, and almost as many worried they would be offered a worse deal. Ralph Tucker, founder of Empath AI, which reads vocal signals to detect distress, puts it plainly:

“The industry is over reliant on self-declaration, and it doesn’t work for a whole number of reasons. What if you’re unaware? What if you’re unwilling? What if you’re talking to a complete stranger?”

If self-identification is the only way in, the gap will never close. Detection has to move into the background, reading how someone engages rather than asking them to fill in a form. Systems can now analyse the tone of a conversation or the pattern of a person’s behaviour over time, and pick up the early signs of strain long before a customer would think to raise them.

Tucker’s emphasis is on prevention: intervening early, rather than waiting for the crisis.

From questionnaires to care

This is where digital money coaching fits. The FCA frames vulnerability around four drivers: health, life events, financial resilience and financial capability. These drivers surface on their own when someone engages with content and tools built around a real life moment, such as buying a first home or losing a job. Money Coach starts from that human context and earns attention by being helpful rather than by selling. Detection becomes a by-product of being useful, the opposite of a tick-box.

We have seen the engagement this creates. My Money Vibe, the Money Coach product we run with Santander, has reached a pilot audience of around 10,000 customers, with confidence rising markedly after use. The same engagement that helps a customer also signals to the firm who might need more. And detection is only the start; the regulator’s next test is proving the support actually improved the outcome. Sophie Legrand-Green of TISA also argues the label itself is part of the problem:

“Vulnerability, like accessibility, is just another flavour of personalisation. We’ve got to get over treating it as a separate thing. It’s about tailoring the experience to the needs of the individual.”

The questions worth asking

So there are a few questions worth putting to the industry. If self-declaration misses most of the people it is meant to catch, how much longer should it be the default? What would it take, in the rules and in practice, to treat everyday engagement as a legitimate route to detection? Where does listening to customers feel like surveillance and what disclosure should there be?

With over 8 years under our belts of designing, developing and testing personalised digital experiences, this is a space we can have an impact on. We are already making good progress building vulnerability detection into our products, with care being a key pillar of our customer trust framework. But there is much more to do. If you are working on this too, we would be keen to compare notes, and we’ll be taking our thinking to the TISA vulnerability working group. Closing the detection gap will take the whole market, because the customers most likely to need help are, almost by definition, the least likely to ask for it.