Agora.io · 2022 to 2024 · Portfolio scope
Designing across six Agora products, each tied to a business outcome.
App Builder, Agent Studio, Console, the marketing site, Performance Lab, and Conversational AI demos. Different surfaces, one platform, one set of business metrics moving in the same direction.
Role
Principal Product Designer
Scope
Six product surfaces
Timeline
2022 to 2024
Focus
CSAT, MRR, activation, brand
Featured · App Builder
Improving CSAT and MRR
Redesigning App Builder around the business metrics that matter.
App Builder is Agora's no-code vertical for integrating video, audio, and live-streaming into an app without touching the SDK. The product worked, but the revenue curve did not. B2B customers churned before converting, pro users hit productivity walls, and trust issues around privacy kept enterprise deals from closing.
Four months of research, redesign, and feature work across conferencing verticals like video calling and live streaming, aimed at one goal: move CSAT and MRR together.
38%
Conversion lift
2×
MRR
↑
CSAT
WCAG
AA passed
01
Lack of trust blocked enterprise deals.
Privacy and security concerns surfaced in every B2B call. Users would not commit until those gaps closed.
02
Subpar UX ate productivity.
Pro users wanted split screen, focus mode, accessible flows. Task completion was slower than competitors.
03
Feature gaps drove churn.
Customers asked for virtual backgrounds, speech-to-text, and engagement tools before they would convert.
Features that moved the metrics.
Three additions closed most of the gap between App Builder and the competitors enterprise buyers were comparing it to.
Accessibility
Speech to text.
Live transcription for users with hearing loss, and for anyone joining from a noisy environment.
Privacy
Virtual background.
Hide the home office, reduce reveal of personal space. A blocker cleared for enterprise buyers.
Brand fit
Magic variable mapping.
Customers edit theme and color tokens inside Agora Console and see them flow into their deployed App Builder app. This is the first bridge between Console and App Builder, the seed of the ecosystem story the rest of this case study is about.
How products connect
App Builder sits in the middle of the funnel, not the end of it.
The CSAT and MRR work on App Builder only paid off when the surrounding products sent the right users in, and handed them off when they outgrew it.
Each surface around it had its own job: attract, convert, onboard, scale, monitor. Below is how they pass users between each other.
Attract
Marketing site
Sends qualified devs to Console
Convert
Demos + Perf Lab
Proof before signup
Onboard
Console
Issues keys, routes to product
Build
App Builder
No-code path, brand tokens
Scale
Agent Studio
When no-code runs out of room
One user, five products. The design job was making sure none of those handoffs felt like starting over.
Across the portfolio
One problem per product, one metric per problem.
The App Builder story above is the deepest one. The table below is the breadth. Each product had its own brief, its own team, and its own number to move.
App Builder
Enterprise churn, CSAT drag, feature gaps vs. competitors.
38% conversion lift, 2× MRR, CSAT up, WCAG AA passed.
Agent Studio
Days to ship a voice AI agent, high drop-off on first try.
Time to first agent: 2+ hours → under 30 minutes.
Console
Friction moving from signup to a working project.
Higher repeat deploys, fewer support tickets on onboarding.
Marketing website
Brand perception lagged the product. Prospects arrived cold.
Improved brand perception, cleaner handoff into Console.
Performance Lab
Model cost and latency data lived far from where devs chose.
Benchmarks surfaced inside Studio config, informed picks.
Conversational AI demos
Prospects could try the tech, but left at a dead end.
Demo sessions now route into Studio signups.
For the deep dive on Agent Studio, see Case Study 01.
Results
What the portfolio moved, together.
Individually each product had its own win. Put together, the numbers describe an ecosystem that is doing what an ecosystem is supposed to do: compound.
38%
App Builder conversion lift
2×
App Builder MRR
30 min
Time to first agent in Studio (was 2+ hrs)
6
Products ladder into one funnel