# Versive — Complete Product Documentation

> Versive is an AI-powered user research and testing platform. Run usability tests with AI personas or AI-moderated surveys with real participants. Get actionable insights in minutes instead of weeks.

Versive is backed by Y Combinator, SOC 2 Type 2 certified, and built by Mirio, Inc.

- Website: https://getversive.com
- Contact: support@getversive.com
- Trust Center: https://trust.oneleet.com/versive?tab=compliance
- LinkedIn: https://www.linkedin.com/company/versivehq

---

## Table of Contents

1. Why Versive
2. Product Overview
3. AI Tests (Testing with AI Users)
4. Studies (Research with Real Users)
5. AI Analysis & Insights
6. Collaboration & Team Features
7. Security & Compliance
8. Pricing & Plans
9. API & Integrations
10. Who Uses Versive

---

## 1. Why Versive

### One Platform for AI and Real Users
Versive is the only platform that combines AI usability testing with AI-moderated real-user research in a single tool. Most competitors focus on one or the other — Versive does both, with shared analysis, personas, and team workspaces. There's no need to switch between products for different research methods.

### Most Comprehensive Question Types
With 18+ question types — including AI conversational questions with intelligent follow-ups, quantitative scales (NPS, rating, matrix, star rating), card sorts, rankings with matrix, allocation, prototype tasks, and media uploads — Versive offers the widest range of any AI research tool. This means researchers can run complex, mixed-method studies without compromising on question design.

### Powerful Logic and Configuration
Advanced conditional branching with multiple operators, screen-out logic, device enforcement, screen recording, custom branding, copy/paste controls, and quota management give researchers full control over every aspect of study design. Most AI survey tools offer basic skip logic at best — Versive supports the kind of complex flows that professional researchers need.

### Built-in Concept and UI Testing
Test Figma prototypes, live websites, and uploaded images directly with AI personas. Choose from dynamic interview mode, heuristic evaluations against Nielsen's 10 heuristics (or your own custom criteria), or structured feedback. Prototype task questions with goal node detection let you measure task completion in real-user studies too.

### Scales from Startup to Enterprise
A free plan lets early-stage teams start testing immediately. Team plans add collaboration features for growing companies. Enterprise plans include SSO, API access, custom limits, and priority support for large organizations. Research agencies use Versive's multi-tenant workspaces and custom branding to run studies for multiple clients from a single account.

---

## 2. Product Overview

Versive provides two complementary product lines on a single platform:

### AI Tests
Run instant usability tests using AI-generated personas. No recruiting, no scheduling, no waiting. Import a Figma prototype, paste a website URL, or upload images, and get a detailed usability report with transcripts, screenshots, and heuristic evaluations — all in minutes.

### Studies
Create AI-moderated surveys and interviews for real user research at scale. The AI interviewer conducts conversational interviews, asks intelligent follow-up questions, and adapts to participant responses. Supports text, voice, and video modes with 18+ question types, conditional logic, and 20 language translations.

Both product lines share a unified analysis layer with AI-powered insights, theme extraction, sentiment analysis, and exportable reports.

---

## 3. AI Tests (Testing with AI Users)

### 3.1 Test Types

#### Figma Prototype Tests
- Import any Figma prototype by pasting a URL
- AI users navigate through interactive prototype screens
- Interaction detection reads Figma's defined interactions
- Select specific frames to include in the test
- Custom screen ordering for controlled test flows
- Goal node specification for task completion tracking

#### Website Tests
- Test live, publicly accessible websites
- AI user operates a real browser
- Task-based testing: define what the user should accomplish
- Step-by-step action log including: navigate, click, type, scroll actions
- Screenshot captured at each step with reasoning for the action
- Configurable viewport sizes:
  - Desktop: 1920x1080, 1536x864, 1366x768, 1280x720, 1024x768
  - Mobile: 500x1080
- Live view: watch the AI user navigate in real time
- Full session video recording
- Task completion status with confidence score and abandonment reasons

#### Image Tests
- Upload static screenshots or design mockups
- Get feedback on visual design, layout, information hierarchy
- Supports multiple images per test

#### Video Tests
- Upload video walkthroughs or product demos
- AI analysis of the video content

### 3.2 AI Personas

Versive lets you create detailed AI personas that simulate real users. Personas are reusable across tests.

#### Persona Builder Fields
- **Summary**: Brief description of the persona
- **Identity**: Demographics, background, role
- **Context**: Situation, environment, constraints
- **Goals & Outcomes**: What the persona is trying to achieve
- **Friction Points**: Known pain points and frustrations
- **Behavioral Patterns**: How the persona typically behaves
- **Interaction Style**:
  - Communication style
  - Technology comfort level
  - Decision-making process
- **Psychometric Profile (Big 5)**:
  - Openness to experience
  - Conscientiousness
  - Extraversion
  - Agreeableness
  - Neuroticism
- **Unique Factors**: Anything else that makes this persona distinctive

#### Persona Generation
- Upload research files (interview transcripts, survey data, user research documents) to auto-generate personas
- AI distills specific details, names, numbers, and quotes from source data
- Content limit: 500,000 characters for rich, detailed personas
- Generate multiple personas from a single research dataset

#### Multi-Persona Testing
- Run the same test with multiple personas simultaneously
- Compare how different user types experience your product
- Each persona execution counts as one credit

### 3.3 Test Modes

#### Dynamic Interview Mode
An AI moderator interviews the AI persona about each screen, producing natural conversational transcripts. Custom interview scripts can be defined per screen.

#### Expert / Heuristic Mode
An AI expert evaluates screens against usability heuristics:
- **Default**: Nielsen's 10 Usability Heuristics
  1. Visibility of system status
  2. Match between system and real world
  3. User control and freedom
  4. Consistency and standards
  5. Error prevention
  6. Recognition rather than recall
  7. Flexibility and efficiency of use
  8. Aesthetic and minimalist design
  9. Help users recognize, diagnose, and recover from errors
  10. Help and documentation
- **Custom heuristics**: Define your own evaluation criteria
- **Severity ratings**: Good, Fair, Poor, Not Applicable
- Detailed issues identified per heuristic

### 3.4 Test Results & Reports

- Responses grouped by user flow and screen
- Prioritized recommendations: P0 (critical), P1 (important), P2 (nice to have)
- Persona attribution linking results to specific AI personas
- Shareable public result URLs
- PDF export with formatted reports
- User feedback collection (1-5 star ratings with comments)

### 3.5 Figma Plugin

"Versive: Instant Usability Tests" — available in the Figma Community.
- Import prototypes directly from your design file
- Select specific pages and frames to test
- OAuth authentication with your Versive account
- One-click test creation without leaving Figma

---

## 4. Studies (Research with Real Users)

### 4.1 Study Types

#### AI-Moderated Surveys (Versive Study)
Conversational surveys where an AI interviewer asks questions, listens to responses, and probes with intelligent follow-ups. Combines the depth of an interview with the scale of a survey.

#### Voice Interviews (Voice Study)
Voice-based AI-moderated interviews:
- Multiple voice options to choose from
- Multi-provider transcription with automatic fallback for reliability
- Domain-specific keyword support for transcription accuracy
- Word-level timing for precise timestamps
- Audio and video recording

#### External Studies
Upload and analyze existing research data:
- Audio files (auto-transcribed)
- Video files (auto-transcribed)
- Text transcripts
- CSV batch imports
- AI-powered insights on imported data

### 4.2 Interview Modes

| Mode | Description |
|------|-------------|
| Text Only | Text-based conversational interface |
| Text and Voice | Hybrid — participants can type or speak |
| Voice Only | Pure voice conversation |
| Voice with Preview | Voice interview with question preview |
| Video | Video recording of participant |
| Voice and Video | Combined voice conversation with video capture |

### 4.3 Question Types (18+)

#### Conversational
- **AI Question**: Open-ended with AI-powered follow-ups. Configurable max follow-ups, min/max response duration, and separate follow-up duration settings. Includes an instructions field to guide the AI interviewer.

#### Input
- **Text Input**: Long-form text responses
- **Short Text**: Brief text responses
- **Email**: Email address capture
- **Number**: Numeric input with optional min/max range validation

#### Choice
- **Multiple Choice**: Single or multi-select. Supports "Other" option, max selections limit, and option randomization.
- **Rating Scale**: Likert-style scale with customizable labels for each point
- **Matrix Rating Scale**: Rate multiple items on the same scale. Supports option randomization and "Not Applicable" option.
- **Star Rating**: Configurable icon (stars, hearts, paws) and max value
- **NPS (Net Promoter Score)**: Standard NPS with configurable range (typically 1-10)

#### Advanced
- **Ranking**: Drag-and-drop or dropdown ranking of options
- **Ranking with Matrix**: Rank options across multiple criteria
- **Card Sort**: Categorization exercise. Options: allow new categories, allow new cards, randomize items.
- **Allocation**: Distribute points across options with configurable total. Can reference previous answer values.
- **Prototype Task**: Figma prototype navigation task with goal node specification, optional recording, and completion tracking.
- **Media Upload**: Image/video file upload from participants. Configurable file type restrictions, max files, and max file size.

#### Informational
- **Message**: Display information, instructions, or media with optional timer
- **Calendar**: External calendar booking integration
- **External Link**: Redirect participant to an external URL

### 4.4 Question Assets

Questions can include rich media:
- **Images**: Inline or fullscreen display
- **Figma Prototypes**: Interactive embeds with frame navigation
- **Videos**: Autoplay and loop options
- **PDFs**: Document display
- Layout options: inline or full-screen
- Sizing: fit or scroll

### 4.5 Logic & Branching

Build complex conditional paths based on participant responses:

**Conditions (operators)**:
- Equals, Includes
- Less than, Less than or equal, Greater than, Greater than or equal
- Count less than
- Always (unconditional)
- Include response in next question

**Actions**:
- Jump to a specific question
- End the interview
- Redirect to a URL
- Screen out the participant

Supports both string and integer comparison types. Multiple conditions can be chained for complex flows.

### 4.6 AI Interviewer

The AI interviewer is a conversational moderator that conducts interviews autonomously:

- **Modes**: Quality mode (best results) or Fast mode (faster, cost-effective)
- **Follow-ups**: Automatic follow-up questions based on participant responses, with configurable depth
- **Rephrasing**: AI can rephrase questions to sound more natural in conversation
- **Custom Prompts**: Fully customizable interviewer personality, guidelines, and behavior
- **Retry Mechanism**: Handles invalid or unclear responses with configurable retries
- **Duration Tracking**: Color-coded timers (green/yellow/red) for response pacing

### 4.7 Language Support

Studies can be auto-translated into 20 languages:

Arabic, Chinese, Dutch, English, Finnish, French, German, Hindi, Indonesian, Italian, Japanese, Korean, Polish, Portuguese, Russian, Spanish, Swedish, Turkish, Ukrainian, Vietnamese

Translated content includes: welcome message, thank you message, screen-out message, and all question text. Participants can select their preferred language.

### 4.8 Study Creation Methods

#### AI-Powered Creation
- Describe your research goals in natural language; AI generates the full study
- Import a Figma prototype and get auto-suggested questions
- Iterative chat-based refinement

#### AI Study Assistant
A chat-based AI agent that can edit your study after creation:
- Add, edit, remove, and reorder questions
- Update study settings and configuration
- Modify logic rules
- Generate interviewer notes
- Structured change tracking with before/after state

#### Manual Creation
- Drag-and-drop question builder
- Question library with templates
- Clone existing questions
- Reorder questions freely

#### Templates
- Pre-built study templates for common research needs
- Organization-specific templates
- Community templates

### 4.9 Study Configuration

- **Welcome Message**: Custom title, body, and CTA button text
- **Thank You Message**: Custom title, body, and optional CTA link
- **Screen Out Message**: Custom title, body, and optional CTA link
- **Status**: Draft, Published, Paused, Archived
- **Quota Limits**: Maximum number of responses
- **Access Policies**: Full access, Edit, Edit survey only, View, Invite only
- **Device Enforcement**: Require desktop, mobile, or allow both
- **Copy/Paste Controls**: Prevent copy/paste in text inputs
- **Custom Branding**: Remove Versive branding, add custom support email
- **Screen Recording**: Tab or full-screen recording (configurable per device)
- **Unique Identifiers**: Custom participant tracking

### 4.10 Launch & Distribution

- **Public Links**: Shareable URL for anyone to take the study
- **Embeddable Widgets**: Embed the study on your website
- **VID Links**: Video Interview Device management for remote testing labs
- **Pause URLs**: Allow participants to pause and resume mid-interview

### 4.11 Participant Recruitment

- **Self-service**: Share a link with your own users
- **Respondent Panel**: Recruit from Versive's global B2C and B2B participant pool (powered by Respondent.io)
  - Screener question support
  - Incentive management
  - Participant status tracking (invited, scheduled, attended, etc.)
  - NDA support
  - Calendar booking integration
  - Direct messaging with participants
- **AI Simulation**: Run simulated interviews with AI personas for testing and validation

---

## 5. AI Analysis & Insights

### 5.1 Interview-Level Analysis

- **Summary**: AI-generated overview of each interview
- **Sentiment Analysis**: Positive, neutral, or negative with reasoning
- **Quality Score**: Rating with explanation
- **Fraud Detection**: Automatic spam and low-quality response detection with reasoning

### 5.2 Study-Level Insights

- **Theme Extraction**: Identify common themes across all responses
- **Per-Question Insights**: Summaries, themes, and sentiment per question with response counts
- **Custom Insight Prompts**: Create custom AI analysis views with your own prompts
- **Insight Types**:
  - Paragraph: Narrative summaries
  - Bullets: Key points lists
  - Structured Insights: Findings with attributed quotes

### 5.3 Quote Attribution

All insights are backed by evidence:
- Quotes linked to specific interviews
- Start/end positions in transcript
- Timestamp references for audio/video
- Video clip IDs for instant playback

### 5.4 Analysis Modes

- **Basic mode**: Faster, cost-effective analysis
- **Pro mode**: Higher quality, deeper analysis

### 5.5 Charts & Visualizations

Auto-generated charts for quantitative question types:
- Rating scales
- Multiple choice
- NPS
- Star ratings
- Number questions

### 5.6 Data Filtering

Filter responses by:
- Question responses (equals, includes, greater than, less than)
- Metadata fields
- Interview status (completed, archived, screened out)
- Simulated vs. real interviews
- Multi-value filters
- Saved filter configurations

### 5.7 Exports

- **PDF Export**: Formatted reports with cover page, executive summary, custom AI insights, per-question pages with charts and data tables. Optimized for large studies (1,000+ responses).
- **CSV Export**: Raw data export sanitized for Excel compatibility
- **Video Clips**: AI-generated or manually created clips from interview recordings with thumbnails

---

## 6. Collaboration & Team Features

### 6.1 Organization Structure

- **Organizations**: Top-level tenant for team management
- **Folders**: Hierarchical organization of studies and projects
- **Study-Level Permissions**: Granular access control per study

### 6.2 Roles

| Level | Roles |
|-------|-------|
| Organization | Owner, Admin, Member |
| Study/Folder | Owner, Editor, Viewer |

### 6.3 Team Features

- Email-based team invitations
- Member management (add/remove)
- Activity tracking
- Contextual comments on questions, welcome messages, thank you messages, and screen-out messages
- Real-time interview observation (monitor live interviews)
- Public sharing tokens for read-only result access

---

## 7. Security & Compliance

- **SOC 2 Type 2 Certified**: Independently audited security controls
- **Encryption**: All data encrypted in transit (TLS) and at rest
- **No AI Training on Customer Data**: Versive never uses customer data to train AI models
- **Multi-Factor Authentication**: MFA support for user accounts
- **Row-Level Security**: Database-level permission enforcement
- **Service Accounts**: Isolated system credentials for internal operations
- **Deployment Regions**: US and EU deployments available
- **Trust Center**: Public compliance documentation at https://trust.oneleet.com/versive

---

## 8. Pricing & Plans

Versive offers a free plan and multiple paid tiers. See https://getversive.com/pricing for current plans and pricing.

### How Credits Work
- Credits are shared across AI tests and real-user studies
- Study creation and AI analysis are unlimited on all plans
- Participant recruiting via Respondent is billed separately (pay as you go)

### Enterprise Features
- Custom limits on tests, personas, and seats
- SSO (Single Sign-On)
- API access and custom integrations
- Priority support and training
- Contact: support@getversive.com

---

## 9. API & Integrations

### Public API (v1)
- Create and update studies programmatically
- Manage questions (add, edit, delete, reorder)
- AI-powered study editing endpoint
- Batch question operations

### Figma Plugin
- Available in the Figma Community
- Direct prototype import
- Frame selection and test creation
- OAuth authentication

### Webhooks
- Real-time event notifications
- Custom integration endpoints
- Configurable event types

### Respondent.io Integration
- Access to millions of B2C and B2B participants
- Screener questions and incentive management
- Participant tracking and messaging
- NDA support

### Slack Integration
- Team notifications for study events

---

## 10. Who Uses Versive

### Startups & Early-Stage Teams
- Start testing for free with AI personas — no budget or recruiting needed
- Validate product ideas and designs in minutes, not weeks
- AI-powered study creation: describe your goals and get a ready-to-run study
- Scale up to real-user research as you grow

### Product & Design Teams
- Instant AI usability tests on Figma prototypes and live websites
- Multi-persona testing to understand different user segments
- Heuristic evaluations for design quality checks
- Prototype task completion tracking
- Live view monitoring of AI sessions
- Figma plugin for one-click test creation from design files

### UX Researchers
- AI-moderated interviews at scale with the most comprehensive question types available
- Voice and video interviews with automatic transcription
- Advanced logic and branching for complex study flows
- Multi-language support for global research (20 languages)
- External data import and analysis for existing research
- Custom analysis prompts and insight views

### Enterprise Organizations
- SOC 2 Type 2 compliance and data encryption
- SSO (Single Sign-On) and multi-factor authentication
- Multi-tenant workspaces with folder organization and role-based access
- API access for programmatic study management
- Custom limits on tests, personas, and seats
- US and EU deployment regions
- Priority support and training

### Research Agencies
- Run studies for multiple clients using separate organizations
- Custom branding: remove Versive branding, add client support email
- Flexible recruitment: use the built-in participant panel or bring your own participants
- PDF and CSV exports for client deliverables
- Webhook integrations for custom workflows

### Developers
- Public API (v1) for programmatic access
- Figma plugin SDK for custom integrations
- Webhook system for real-time event notifications
- Embeddable study widgets

---

## Summary

Versive replaces the traditional weeks-long research process with instant, AI-powered testing and analysis. Whether you need quick design validation with AI personas or in-depth interviews with real users, Versive provides the tools to get actionable insights faster than any other platform.

**Get started free**: https://getversive.com/signup
**Request a demo**: https://form.typeform.com/to/CMhju24g
**Contact**: support@getversive.com