Driver History

When someone shops for insurance, a third-party system checks their driving history to validate what they entered, and that check directly shapes the final quote. The old experience gave people no window into any of it. No explanation, no context, no moment of understanding.

The business came to UX with a clear ask: surface accident and ticket capture on the front end, so customers could see what was going into their price. The reasoning was practical. Insurance is expensive enough in some regions that transparency at the right moment builds trust. This started as a request to add a step to the quote, right between adding drivers and adding vehicles, not as a redesign of something that already existed. My job was making that added step feel genuinely useful, not punitive.

UX lead. Before sketching anything, I mapped the backend logic to understand exactly what the system was doing and where it could create problems for people. From there I led a cross-functional kickoff with product, research, UX, and content to align on goals and constraints early. One question guided the whole project: how do you add transparency without adding friction?

Low-fidelity wireframes explored how to capture accident and ticket information in context, without turning the step into a roadblock. The iterations focused on a few questions at once: where in the flow the step should live, which interaction patterns felt least clinical, and how the copy could reframe the moment.

I tried dropdowns against free text against toggles, weighing which one carried the disclosure without feeling like an interrogation. And I kept pushing the language toward something that read as help rather than an audit, since the words in this step directly shape how honestly people report their history.

Research testing with ten participants surfaced three specific friction points, and each one shaped the final direction.

Deletion was not intuitive. Half the participants struggled to correct an incident after deleting it, because the corrective action was buried in a dropdown, the warning was too quiet, and the save button was easy to miss, so some people hit back instead of finishing the fix. The close icon was ambiguous, with two in ten unsure whether the X closed the dropdown or deleted the incident entirely, a small thing with real consequences in a high-stakes flow. And the language was doing work the design was not supporting: most people understood "accident" right away but were unsure about "incident" and "violation," treating incident and accident as the same and violation as loosely defined.

The language finding had a direct fix. A tap on the info icon defines what actually counts as an accident or a ticket, in plain language rather than policy jargon, so people learn what qualifies before they answer. It defuses the "will this hurt my quote?" worry at the moment it spikes.

The step itself threads into the flow in three beats. It starts as a single yes or no question inside Add Drivers, so only drivers with history go any further and nothing extra is asked of people it does not apply to. When the answer is yes, details appear one incident at a time, with plain language guiding recall instead of insurance terminology. A summary card then confirms each entry, with the option to add another or continue the quote. The result is a legal disclosure that reads as transparent rather than punitive.