Smart Vehicle Lookup
Component Design & Validation
Solving a fundamental knowledge gap — most drivers don't know their exact vehicle trim. A tag-based dynamic filter lets users browse to their car rather than recall precise specs, cutting early-stage abandonment significantly.
The Core Insight
Most people don't know their car's exact trim
Through user research with our UX research vendor, we discovered a fundamental assumption failure: the auto quote journey assumed users knew their vehicle's exact trim, transmission type, and drivetrain. In reality, most don't — especially new drivers and those who purchased secondhand.
The existing typeahead required a precise string: "2021 Honda Accord Sport 1.5T FWD CVT" — and failed if any part was wrong. This caused abandonment at Step 1 before users ever saw a price.
"Why are we asking people to recall information they've never had to know?"
Research sessions: users saying "I know it's a Honda Accord from 2021, but I have no idea what trim." The solution wasn't to educate users on vehicle specs — it was to design around that knowledge gap entirely.
The Problem
A typeahead that punished normal users
Single Precise Entry Required
Required users to type their exact vehicle in one field. Any mismatch = zero results or wrong options. No recovery path.
Knowledge Gap Assumed Away
Most users know year + make + model — not trim, transmission, or drivetrain. The form demanded expert knowledge to proceed.
Especially Painful on Mobile
Typing a long precise string on mobile with autocorrect was a friction multiplier. Drop-off was highest on mobile devices.
❌ Old — Precise String Required
✓ New — Tag-Based Smart Filter (My Design)
Smart Vehicle Look-up
Showing 6 matches — add "Sport" to narrow to 2
"Entering '2021 Honda Accord' showed 6 models. Adding 'Sport' narrowed it to 2 — and users recognised their car immediately."
— Usability test observation, Surbhi Mahendru
The Solution
Browse, don't recall — a component that meets users where they are
Individual Tags, Not a Single String
Each known detail (year, make, model, trim) is added as a separate tag. Users add only what they know — the system narrows results with each tag. No single point of failure.
Real-Time Dynamic Narrowing
As each tag is added, results update instantly. "2021 + Honda + Accord" shows 6 models. Adding "Sport" narrows to 2. Users browse to their vehicle rather than recall it perfectly.
Contextual Sub-Labels
Each result shows plain-language details: "4-cyl · Automatic · Front-wheel drive." Users identify their car by description, not by technical spec string alone.
The Mobile Insight — Contextual Responsive Design
Post-launch usability testing revealed mobile users preferred a three-dropdown format — similar to competitors. Managing tags on a small screen had higher cognitive load than anticipated. This led to a responsive adaptation: smart tag filter on tablet and desktop, traditional three-dropdown on mobile — same database, two presentations. Design is never finished.
Impact & Learnings
What the data and testing revealed
Early Completion Rate
Smart tag filter significantly improved Step 1 completion for desktop and tablet users post-launch.
Results from 4 Tags
"2021 + Honda + Accord + Sport" = 2 precise results. Eliminates guesswork for most vehicles entirely.
Mobile Insight Unlocked
Post-launch testing revealed mobile users preferred dropdowns — leading to a responsive variant strategy.
Context Changes Everything
The same component that delighted desktop users frustrated mobile users. Patterns aren't universal — they're contextual. Test in the environment your users will actually use the product.
Browse vs. Recall is a Fundamental UX Decision
Many forms ask users to recall information they don't have. Switching to browse patterns — especially for complex product databases — is often the highest-impact change you can make.
Launch is Not the Finish Line
The mobile insight only emerged post-launch. Building in a usability testing cycle after launch — not just before — is essential. This component went through three iterations in six months.
Final Reflection
"Design is never finished — each insight shapes a more intuitive experience."
The Smart Vehicle Lookup proved that a single well-designed component can meaningfully change a product's funnel performance. But it also taught me that "correct" design isn't static — it's responsive to context, device, and evolving user behaviour. The most valuable outcome wasn't the completion rate improvement — it was the usability insight about mobile patterns that shaped future component decisions across the Aviva design system.