Micro-frictions in banking and insurance journeys

Two in three people who start opening a bank account online never finish. That is not a typo. Signicat’s Battle to Onboard research, run across seven European markets, put digital banking abandonment at 63% in 2020 and 68% by 2022, up from 40% in 2016. The trend is moving the wrong way, and it is costing the industry real money: one modelling exercise estimated 5.7 billion euros wasted every year in Europe on abandoned financial onboarding alone.

Here is what makes it interesting. These are not tyre-kickers. Someone who clicks "Open an account" has already decided they want the product. They arrive with intent, and they leave anyway. The reason is rarely one big wall. It is a series of small ones.

I call them micro-frictions: individually minor, collectively fatal. A field that asks for information the user does not have to hand. A form that never explains why it needs their ID, or what happens to it. A step that gives no sense of how much is left. A phone-number field that rejects a valid format. None of these alone would stop a motivated person. Stacked together, at a moment when the user is already nervous about handing over sensitive data, they add up to "I’ll do this later", and later never comes.

Regulated sectors make this worse, because friction is partly designed in. KYC and anti-money-laundering rules force steps a retailer would never impose. That is not optional. But teams too often treat compliance as a licence to stop caring about the experience around it. Banks ask for more than they strictly need, hoping to populate a CRM or pre-empt a future requirement, what one industry writer neatly called "questionnaire creep". Baymard’s checkout research points the same way: trimming a form from the industry-average of roughly 23 fields to the 12 to 14 that are actually necessary can lift completion by about a third. Every extra field is another chance to lose someone who was ready to say yes.

Insurance runs on the same logic. A quote form is a trust negotiation. The user is being asked to hand over personal details and, eventually, money, for a promise they cannot see or touch. When the form is long, the value unclear, or the reassurance thin, they bounce to a comparison site and never return. The product might be excellent. The journey lost them before they found out.

The uncomfortable part is that most of this is invisible in a standard dashboard. Aggregate conversion tells you people are leaving. It does not tell you they abandoned on the identity step because the copy never explained why a passport was needed. That level of diagnosis needs step-by-step funnel analysis and session replays: watching where the hesitation actually happens, not guessing from a topline number.

And the diagnosis usually points to confidence, not effort. In a regulated journey, users do not abandon because a form is long. They abandon because, at the highest-stakes moment, nothing told them they were safe to continue. Signicat found that a bad onboarding experience makes more than half of customers less likely to use that bank at all, and a third warn their friends off. The friction does not just cost the conversion. It costs the relationship.

Fixing it is rarely a redesign. It is finding the three or four moments where trust quietly breaks, and repairing those: say why you need the data, show progress, cut the fields that serve you and not the user, add the reassurance a nervous person is looking for. Small moments. They are where regulated journeys are won and lost.