Prototype testUnmoderated

A/B prototype testing template

Compare two versions of a web prototype with real users. Participants take a 5-second first-impression test, try both designs in random order while thinking aloud, then rate ease and pick a favorite.

No sign-up or credit card needed.

Tasks & questions
12
Time to complete
~10–12 min
Suggested sample
8–10 participants
What's in this study
12 tasks and questions in 4 sections
Open in editor
About the participant
When did you last use a website or app to plan a leisure trip?
Which of these have you used to plan a trip in the past 12 months?
Rank what matters most when planning a leisure trip
First impression
View the homepage for five seconds
What do you remember most about the page?
Design A vs. B (everyone sees both, in random order)
Plan a trip in the live prototype while thinking aloud
How easy was it to find a hotel and reach the booking review?
What stood out most about this version?
Wrap-up
Which version would you prefer to use?
What would your ideal experience look like?
Where would this fit in how you normally plan a trip?
Is there anything else you'd like to share?
When to use it

Use this template when you need to pick a direction

01

Choosing between two concepts

Two layouts, flows or navigation models? See which one people understand and complete more easily before engineering commits.

02

Testing a redesign against today

Put your current experience up against the new one to show whether the change actually makes the task easier.

03

Settling a design debate with evidence

Replace opinions with task ease scores, screen recordings and a clear preference split your stakeholders can see.

FAQ

Prototype testing FAQs

What is a prototype A/B test?

A prototype A/B test compares two versions of a design with real users before you build it. Participants complete the same task on each version while thinking aloud, then rate how easy it was and say which they prefer, so you learn which design works better and why.

How many participants do I need for a prototype A/B test?

Because everyone tries both designs, 8–10 participants usually surface the main usability issues and a clear preference. Recruit 20–30 if you want differences in ease scores that you can report with more confidence.

Should each participant see both designs?

This template has everyone try both versions, and Hubble shuffles which one each person sees first so order effects even out. If the tasks are long or the designs are nearly identical, you can switch to showing each person only one version.

Do I need a Figma file?

No. The prototype tasks open any URL, so you can test a coded prototype, a staging site or a live page. Figma prototypes work too.

How long does the study take?

About 10–12 minutes per participant: three screener questions, a 5-second first-impression test, a think-aloud task and two follow-ups for each design, and four wrap-up questions.

Who can I recruit?

Share the study link with your own users or customers, or recruit from Hubble's participant panel and filter by demographics, habits and devices.

Can I change the tasks and questions?

Yes. Open the template in the demo editor to edit, add or remove any task or question, then swap in your own prototype URLs.

Find out which design works better

Open it in the demo editor and make it yours. No sign-up needed.