Percura
Percura / ProductSelf-updating

Every test you run makes the next one more accurate.

The intelligence layer behind Percura. A living graph of connections between persona types, friction patterns, UX failure points, and real-world behavioral signals — built from millions of data points and updated continuously.

3M
Indian persona profiles mapped
20+
Behavioral archetypes in the graph
6
App categories indexed
Weekly
Graph updates from live signals
01Graph Structure

What the Graph Connects

Real-world signals
Play Store reviewsReddit complaintsSocial signalsValidation feedback
Behavioral archetypes
urbanilliterateinformalmetrograduatetech
Demographic segments
tier-2secondaryblue-collar
Friction patterns
payment drop-offOTP abandonmenttrust collapseform fatigue
Simulation outputs
Simulation parameterFriction reportArchetype update
Real-world signals
Behavioral archetypes
Friction patterns
Simulation outputs
02Maintenance

How the Graph is Built and Maintained

01

Real-world data ingestion

automated · weekly

Every week, the graph ingests thousands of app store reviews, Reddit posts, and social media signals from real Indian users. An AI extraction pipeline turns each one into a structured behavioral signal — which stage failed, what emotion was triggered, who wrote it.

02

Archetype behavioral mapping

20–30 archetypes

Each signal is tagged to one of 20+ behavioral archetypes — groups of personas sharing literacy level, device type, income bracket, and region. The graph learns which archetypes are most vulnerable to which friction types, and how strongly.

03

Validation feedback loop

self-improving

When founders run real user tests after using Percura, they can upload the results. The graph compares AI predictions against real outcomes. Every match strengthens the connection. Every miss triggers a parameter update.

04

Live social signal updates

real-time drift detection

User trust and behavior shifts with the news cycle. A fraud scandal makes a whole demographic more suspicious. The graph picks this up weekly through public social signals and updates archetype parameters accordingly.

03Capabilities

What the Knowledge Graph Powers

Smarter persona selection

When you submit a product flow, the graph recommends which archetypes are most likely to encounter friction. A fintech onboarding flow automatically gets low-trust, payment-sensitive personas prioritized.

Pre-loaded friction patterns

The graph knows that payment screens fail low-literacy users at step 3, OTP requests lose informal workers in 8 seconds, and English-only UIs cause drop-offs in tier-2 cities. Your simulation starts with this knowledge.

Accuracy that compounds

Every simulation run, every customer validation, every weekly signal update makes the graph richer. Our accuracy improves with every test, across every customer. Competitors who start later cannot catch up easily.

04Design Principle

"The graph doesn't forget. Every founder who tests with Percura makes it smarter for the next one."

— Percura Knowledge Graph design principle

Learn how we source our data on the About Us page.