B2C · 2025 · Electronics · Client work
Fewer fields, same data: redesigning a registration funnel
asking for less, getting more
Far fewer products were being registered than sold, and most people quit on step one. I cut step one from five fields to two, without losing any of the data the business needed.
- Role
- UX/UI Designer, discovery to delivery
- Client
- A global consumer electronics brand, US market (anonymised)
- Team
- Me (UX: flow, logic and sequencing) and a designer colleague (component library). Client contact through an intermediary.
- Delivered
- Full registration flow in Figma, plus an HTML build handed to the client
The problem
Most people abandoned registration on the first screen, but the business wanted more data, not less.
What I did
Found the drop-off with reviews and traffic data, agreed one priority with stakeholders, and moved everything but email and model number out of step 1.
How it’ll be measured
Completed registrations, step 1 completion, and emails captured. The client hasn’t shared results yet.
Before · step 1
- First name
- Last name
- Phone
- Address
After · step 1
- Model number
Part 1 of 4
Context
Customers were giving up on registration, and the business wanted to ask them for even more.
The problem
more data, from a form people were quitting
In the US, the gap between products sold and products registered was large. The business wanted more registrations and more customer data, especially email addresses for marketing and Protection Plan sales.
The existing form asked for a lot, up front, and stakeholders didn’t want to collect any less.
The challenge: get more people through registration while still collecting everything the business needed.
Working without direct access
annotate everything
We could only reach the client through an intermediary, so I annotated every design decision in Figma and built presentations explaining the reasoning. Writing down why kept every decision clear, even when I wasn’t in the room.
Part 2 of 4
Research
No budget for interviews, so I went where customers were already complaining, then agreed one clear rule with stakeholders.
Finding where people dropped off
no budget? go where users already talk
I collected user-reported issues from Reddit and Trustpilot, and used website traffic data to see where people left. Three problems stood out:
- Most drop-off happened on step 1, the very first screen
- The model number lookup often failed, so people couldn’t finish
- There was no clear confirmation, so people weren’t sure it had worked
The old flow asked for 14 pieces of information across three steps, and five personal details came before the customer had even said which product they owned.
| Step | What the old form asked for |
|---|---|
| Step 1 | First name, last name, email, phone, address |
| Step 2 | Model number, purchase date, delivery date |
| Step 3 | Serial number, price paid, proof of purchase, buying method, retailer, marketing opt-in |
Agreeing what mattered most
one rule: protect the email
Instead of debating every field, I sent stakeholders a short questionnaire about their main goal and the data that mattered most. The answers: more registrations and Protection Plan sales, and above all, email addresses for marketing.
That gave me one rule for every decision: protect the email address, and make everything else as easy as possible.
Part 3 of 4
Design decisions
Four changes, each tested against the rule: protect the email, make everything else easier.
Step 1 asks for two things, not five
5 fields down to 2
Step 1 now asks only for the email address and the model number, so customers know straight away they’re in the right place. Data is saved as soon as they select Next, so the business keeps the email even if someone leaves later.
A short benefits section on the same screen explains what registering gets you, giving people a reason to start and finish.
Verify later
ask for less, get more
Detailed personal and purchase information became optional. Customers can complete it later from their account, which removes the heaviest part of the form for people in a hurry.
The client wanted more data, and this asked for less up front, so it wasn’t an easy sell. I showed how it served their own goal: people who come back to finish are more likely to complete it than someone facing a long form on day one. They approved it.
Pushing back on a dense layout
evidence over taste
Stakeholders wanted the form visible without scrolling, and their first instinct was a dense two-column layout that matched the brand’s home-market conventions.
I argued that US customers expect a lighter, step-by-step form, and that asking for a lot up front was exactly what caused drop-off. I backed this with accessibility evidence.
Result: a single-column form with two fields visible without scrolling.
Fixing what broke registration
- Model number: the rigid dropdown became a search field that suggests models as you type, with guidance on where to find the number on the product.
- Confirmation: a clear confirmation email plus a print option, so customers always have proof it worked.
- Protection Plan: offered at the final step, making clear it’s provided by a third party.
Part 4 of 4
Outcome
The flow was handed over and is being built. Here’s how I’d judge whether it worked.
How success would be measured
no data yet, so here’s how I’d judge it
I handed over the full flow in Figma and an HTML build. We heard the client’s developers were implementing it, but despite asking, we haven’t received data yet. These are the measures I proposed:
| Goal | Metric |
|---|---|
| More registrations (primary) | Number of completed registrations |
| Fewer people quitting early | Step 1 completion rate compared with the old flow, and drop-off between email capture and verify later |
| Marketing data | Emails captured, including from people who don’t finish |
| Protection Plan sales | Plan sign-ups from registration |
What I learned
Agreeing priorities with stakeholders up front made every later decision easier. When disagreements came up, I could point back to the goals we’d already agreed.
In-house, I’d A/B test the new step 1 against the old one, track completion at every step, and use the results to improve steps 2 and 3.