Compare B2B Account Segments in Amplitude
Use account-level reporting when the question concerns companies or teams rather than individual people. Verify the Accounts capability and group instrumentation first; a user property containing a company name is not automatically equivalent to group analysis.
Before you start: Have the relevant account-level capability, stable group identifiers and verified group attribution. Agree whether a conversion can be completed by multiple members.
Worked example: Compare active paid workspaces by region. One member may create a project while another shares it, so account conversion can differ from a funnel requiring the same person to do both.
Step-by-step guide
Define the account
Choose a stable group type and identifier. Avoid display names that can change or collide.
Verify instrumentation
Inspect events and membership to confirm attribution to the intended account. Check the difference between event-level and persistent user-level grouping.
Choose account outcomes
Define activation, adoption or conversion at the group level. State whether several members can contribute to the same outcome.
Build comparable segments
Use relevant account properties and equal observation windows. Keep large and small account composition visible.
Inspect sample accounts
Review known successful and unsuccessful groups. Verify that cross-account users do not contaminate attribution.
Record limitations
Document group coverage dates and reporting capability assumptions. Do not imply historical group attribution exists before it was instrumented.
Common mistakes
Company name equals group setup
A label alone does not configure account reporting.
User and account funnels mixed
Their completion semantics differ.
Historical attribution assumed
New group instrumentation may not rewrite the past.
Tips
Use immutable account identifiers.
Inspect a person belonging to multiple accounts.
Report account size alongside segment outcomes.
Sources and verification
Based on product documentation checked on 2026-10-24. This is a documentation-based guide, not a hands-on benchmark. Features and availability can change.