User Testing & Bug Reports
Embed a Magnet widget in your app. Users click, record their screen + voice showing the issue, and submit. AI transcribes, detects frustration level, and auto-creates tickets via webhook.
By eliminating back-and-forth "which button?" emails
Magnet
Video Forms14-day trial · No credit card
Sound familiar?
“Screenshots miss context. Text descriptions miss steps. You're playing detective instead of fixing bugs.”
Step-by-Step Guide
How does it work?
MagnetAdd Magnet widget to your app
Paste one line of JavaScript into your app. A small floating button appears — users click it to report issues instantly without navigating away from the page.
Choose widget shape: circle button, vertical side tab, horizontal bar, or popup — match your app's UI
Target by URL pattern: show only on /checkout, /settings, /dashboard — where bugs happen most
Device targeting: show on desktop only (screen recording works best on desktop) or all devices
Optional greeting message with delay: "Having an issue? Click here to show us" appears after 5 seconds on the page
Enable scroll trigger at 80% scroll depth on long pages — users who scroll deep and encounter issues are most likely to report. Also enable exit intent to catch frustrated users before they leave.
MagnetUser records screen + narration
User clicks the Magnet button, grants screen share, and narrates the issue in real-time. The entire context is captured in one recording — no 5-email back-and-forth.
"I click checkout, enter my card number, hit submit, and it just spins forever" — you see the exact URL, browser, button, and visual state
Device info auto-captured with every submission: OS, browser version, screen resolution, country — no need to ask
Users can also upload screenshots or files alongside the recording for additional evidence
AI auto-prioritizes by frustration level
AI processes every recording through the full analysis pipeline — transcription, sentiment, emotion detection, keyword extraction. Frustrated user = urgent. Neutral = backlog.
Sentiment breakdown: positive % / negative % / neutral % — with specific emotion: frustrated, angry, sad, neutral, satisfied
Time-stamped transcript: click any segment to jump to that moment in the video — engineers find the bug in seconds
Keyword extraction surfaces: "checkout", "payment", "spinning", "stuck" — searchable across all bug reports
AI Config: configure specific analysis rules — "Flag any mention of data loss, security concerns, or payment failures as P0"
Auto-create tickets with full context
Webhook sends the complete package to your project management tool. Engineers get everything they need in one ticket — zero context-switching.
Webhook payload: video recording URL, full transcript, sentiment score, emotion label, device info, page URL, user email
Conditional routing: negative sentiment + "payment" keyword → P1 ticket in Jira; neutral sentiment → feature backlog in Linear
Email notification to the engineering team lead when a critical bug is detected — real-time alerting
Use webhook with Bearer auth and retries (5 attempts, exponential backoff) to ensure no bug report is lost — even if Jira is temporarily down.
Connect with your existing tools
Automatically route responses to your CRM, project management, and communication tools.