Used For

• +**Your checkout funnel is hemorrhaging users somewhere between cart and payment** → See exactly which step loses 60% of people and watch session replays of them rage-clicking the broken button
• +**You're running A/B tests blind, hoping the green button beats the blue one** → Test changes on 10% of users, see real behavior data, roll out winners automatically
• +**SQL queries for user cohorts take your analyst 2 days** → Drag-and-drop charts give you the same insights in 30 seconds
• +Behavioral cohorts automatically group users by actions - target "viewed pricing but didn't buy" without writing complex queries
• +Session Replay shows you the exact moment users give up - like watching over their shoulder as they try to use your app

Best For

• >Your app's retention tanked 15% and you have no clue which feature broke
• >Manual funnel analysis in spreadsheets is eating 20 hours of your week
• >Hit 50K users and Google Analytics feels like tracking ants with a magnifying glass

Not For

• -Solo founders or teams under 10 people — you'll hit the 10M event limit in weeks and face surprise $600 bills
• -Anyone wanting simple website traffic data — this is overkill if Google Analytics does what you need
• -Companies under 5K monthly users — the setup time won't pay off with so little data to analyze

Pairs With

• *Snowflake (where all your customer data lives so you can connect behavior to revenue)
• *Segment (to collect events from your app without coding tracking into every button)
• *Slack (where your PM gets alerts when conversion rates tank)
• *Mixpanel (what you're probably migrating from because Amplitude plays better with data warehouses)
• *dbt (to transform raw event data into something your executives can actually understand)
• *Salesforce (to see which user behaviors predict who becomes a paying customer)

The Catch

• !The "generous" 10M event free tier fills up fast - apps with 5K daily users burn through it in weeks
• !You'll need someone to spend 1-2 hours weekly maintaining event schemas or 20% of your data goes missing
• !Session Replay and advanced features add compute costs that aren't obvious upfront - enterprise support is great but onboarding consultants tack on $5K

Bottom Line

Turns user behavior into charts that don't require a data science degree to read.