Research
/
2024–2025
Meta for Business: Support Panel Usability Testing
Six usability studies over 18 months that took a new support panel for small businesses from early concept to shipped.
Project details
Company
Meta for Business
Role
Research lead, end to end
Methods
6 unmoderated usability studies (82 participants)
Year
2024–2025
Overview
The Problem: Small business owners had to hunt across Meta’s business site for urgent alerts, tips, account info, and expert help. Meta was building one panel to bring it all together, and it had to be instantly clear.
The Method: Six unmoderated usability studies on UserTesting and dscout, each building on the last: from static images, to video concepts, to a clickable mobile prototype.
The Core Insight: People could always tell an urgent alert from a tip. They struggled to find them and to tell how many there were. Explicit beat subtle every time.
The Impact: The stacked notification pattern and my final recommendations shipped in Meta for Business.
Team: I led all of the research, partnering with designer Ray Chang, Meta’s creative direction team, and the agency Code and Theory, who built the prototype.
What was this?
A new support panel for small businesses on Meta’s business website.
When small business owners are on Meta’s business site, the tools they need are scattered: urgent alerts about failed payments or policy violations, tips for better results, their account details, switching between businesses, and access to Meta Pro experts. Meta was building one panel to bring all of it together, so business owners could see what needs attention and act on it in one place. I led the usability research on that panel from its earliest concepts to the version that shipped.
How did I test it?
Six usability studies, raising the fidelity each time.
Each round answered one question with the cheapest test that could answer it. I started with static images, moved to video concepts, then a clickable mobile prototype. Every study used the same backbone: a realistic task, think-aloud on a screen recording, 1–5 ratings for findability, clarity, and ease, and open follow-up questions. When I compared designs, I rotated the order so no version won just by going first.
Can people tell urgent alerts from tips?
Study 1: Comprehension was solved. Findability wasn’t.
Tested: 10 people classified 6 notifications, then pointed to where they’d find them on two page designs.
Found: 60 of 60 classifications were correct, with the visual cues rated 4.7/5. A red triangle and “Resolve” meant urgent; “Recommended for you” meant a tip. Findability confidence was 3.8/5 with a red dot on a chat button, and 4.6/5 with a bell and a count. 9 of 10 preferred the bell.
Changed next: we kept the visual cues and turned to where notifications should live.
Where should notifications live?
Studies 2–4: Three fast tests on placement and navigation.
Under the profile photo: easy to spot (4.1/5), but comprehension fell to 3.5/5. People read it as tips, and urgent alerts got buried under “Support.” → Moved notifications to their own entry point.
Tabs vs. icons with counts (A/B, order rotated): icons with counts won, with finding tech support at 4.8 vs. 4.4 and finding tips at 4.7 vs. 4.1. → Added a count for each type.
Arrows vs. swiping: arrows and dots edged ahead, 6 of 10. → The margin was thin, so I tested it properly next.
How do you show there’s more than one?
Study 5: Explicit beats subtle.
Tested: 24 people watched three layouts (carousel, side-by-side, stacked) in rotated order across three groups, then ranked them.
Found: stacked won 2 of 3 groups because “See 3 more” spelled it out. Carousel dots read as one message repeating, and side-by-side cards “just slid by real fast.” People wanted to see the whole set before acting on any one.
Changed next: stacked went into the prototype.
Does it hold up once it’s built?
Study 6: Five features, one clickable prototype.
Tested: 18 small business owners completed five task-based tests on mobile, with think-aloud and 1–5 ratings for clarity, findability, and ease.
Found: the notification pattern worked. “5 new alerts” and “See all notifications” did their job. The usability issues had moved elsewhere:
Red alerts read as urgent instantly, but tips below them often didn’t register as notifications.
The account page read as a brochure, not a profile. Only 12 of 18 could tell which business it was for.
16 of 18 found the account switcher, mostly out of Instagram habit, and wanted confirmation after switching.
Meta Pro read as premium human support, but people couldn’t tell if it was free.
What did I recommend?
Say it plainly.
Ship the stacked pattern with explicit counts like “See 3 more”
Frame tips so they read as part of the notification system
Relabel the account page around what it actually is, and show the business name clearly
Add an account switcher at the top, with a “You’ve switched to…” confirmation
Say plainly whether Meta Pro is free and whether you’ll reach a real person
Where does it fall short?
I was upfront about its limits.
Early rounds used static images and videos, which test how people read a design, not real use. That’s why the last round moved to a clickable prototype.
About 8 people per group in Study 5, so the ranking is directional. The written feedback was consistent.
Study 6 covered five features in one session, which traded depth for breadth.
What came of it?
Outcome
Set the UX direction for the panel’s notifications and account experience
The stacked notification pattern and my final recommendations shipped in Meta for Business
Took the panel from early concept to shipped in six studies over 18 months