Variant A
Dropped
Dense form
The status quo. Ten-plus fields up front, everything asked at once, before anyone reached an advisor.
- All fields on one screen
- Reads like homework
- Drop-off above the fold
Scope
Finanzcheck is Germany's leading independent comparison portal for installment credit and private financial products. The platform sits between consumers and lenders, a regulated funnel that has to clear strict legal, financial, and consumer-protection requirements while still converting. Conversion in this space isn't optimization theatre. It's the difference between people getting credit decisions in minutes versus abandoning the form three steps in.
The company's priority at the time was unambiguous: produce results. Side projects were cut. The work that lived close to revenue was the work that got resourced.
The remit was the landing-page conversion for the regulated credit funnel.
The credit funnel, the version that shipped was the slimmer one
The brief was direct: redesign the credit funnel. The existing funnel was the kind that converts despite itself, long, dense, asking too much of the user too early.
Interviews, the advisors' feedback and the analytics pointed the same way: the form itself was where people dropped off. Ten-plus fields on one screen read like homework, and the dread of it cost conversions before anyone reached an advisor. The hypothesis was simple, if filling it felt like a conversation instead of homework, fewer people would quit. I owned the UX and UI and aligned the approach with the PO and with Google.
The redesign approach was conversational. Question by question, one screen at a time, with a progress bar at the top. Anything that could be answered later in the advisor call was moved off the funnel to lighten what the user had to do upfront.
The roadmap had more on it than the redesign as shipped. Auto-extraction from ID photos, payslip photos, and bank statement photos, designed to reduce friction further by letting the funnel pull data from documents the user already had on their phone. We wanted those features. We had designs. We knew they'd help.
We didn't get to ship them. The workload was real, time-to-market mattered, and the call was made: ship the slimmed conversational funnel first, sequence the photo-upload features for the next iteration. It was the right call. The slimmed funnel performed, a double-digit conversion uplift, validated through A/B testing in production traffic. The photo-upload features stayed on the roadmap with a clear case for them, made stronger by what the slimmed funnel proved.
Before
One dense screen
Ten-plus fields visible at once, asking too much, too early.
All fields up front
Reads like homework
Cognitive overload above the fold, drop-off below it.
Cost conversions
After
One question per screen
Question by question, progress bar at the top, like a conversation.
Conversational funnel
Lighter upfront ask
Anything answerable in the advisor call moved off the funnel.
Double-digit uplift, A/B validated
Credit funnel, before and after. The original asked everything up front. The redesign asked one question at a time, with progress visible, and moved anything that could be answered in the advisor call off the funnel itself.
With Google Germany
I led the functional UX on the funnel, in direct collaboration with Google Germany.
My side of that partnership was the funnel’s UX and the experiment discipline behind it:
- Mobile-first funnel structure, meaningful, not cosmetic, given the funnel ran majority-mobile
- Form ergonomics on mobile, input modes, autocomplete, keyboard types, smart defaults, conditional logic, pre-validation
- Conversion-coded copy, paired with a UX copywriter and tested per variant
- A/B testing discipline, aligned with Google’s controlled-experiment standard
The conversion architecture, social proof, trust badges, comparison structures, the visual hierarchy of the funnel itself, came from the design team's work.
Conversion design, what came from the team
Beyond Google's bar on performance, the conversion architecture itself came from the design team's work. A non-exhaustive list of what landed in the funnel and on the landing pages:
- Customer testimonials with photos and names
- Trust badges, TÜV, Stiftung Warentest, partner banks, certification logos
- Risk-reversal language, 100% free, no obligation
- Comparison tables that showed what we offered against the alternatives
- Calculator widgets that personalised the offer before the user committed to the funnel
- Estimated time-to-decision callouts, An answer in 2 minutes
- Inline help and tooltips for the fields users hesitate on
- Plain-language explanations for regulated financial terms
None of these are dramatic on their own. Together they compound.
Reflection
Sometimes the version that ships is the one without the features you wanted most. The photo-upload work was the right idea and the wrong moment. Cutting it to ship the slimmed funnel was the right call, and the funnel's performance made the case for the photo-upload features stronger than the original pitch could have.
Conversion in regulated financial products is real product work, not optimization theatre. Every variant tested has to clear legal and consumer-protection requirements before it goes live. The funnel doesn't get fast wins, it gets disciplined wins, taken under constraint.
User research uncovered a blind spot. It identified that consumer credit literacy itself was the friction the platform could remove.
The shape of the work. Landing-page conversion delivered through a cross-functional team in agile cadence. Continuous discovery (user interviews, A/B testing, ideation paired with the UX writer) ran in parallel with delivery. Roadmap definition and feature prioritization made under time-to-market pressure: photo-upload features sequenced for next iteration, slimmer conversational funnel shipped first. Controlled experiments validated each move in production traffic. Freelance designers were brought onto the engagement for the build-out.