Migrate from Certainly to Intercom

Certainly Flows define dialogue trees with conditional branches, fallback policies, and handoff triggers. Intercom Resolution Bots use a step-based builder without direct conditional branching logic. We map each Certainly Flow to an Intercom Resolution Bot, flagging any multi-branch decision trees that require decomposition into multiple bot steps or Fin AI Agent topic routing. Handoff rules to Zendesk agents do not transfer because Intercom does not have a native Zendesk connector; we document the handoff logic for manual rebuild as Intercom assignment rules or team routing.

Certainly intent names and training utterances export as structured data. Intercom Fin AI Agent uses topic-based classification rather than intent-based NLU. We export the full utterance set so your team can create matching Topics in Intercom and train Fin on the same language patterns. We explicitly note that intent classification accuracy will degrade without retraining; Fin requires a Knowledge Hub with representative articles to achieve equivalent deflection rates.

Certainly custom entities and slot types (including regex-based and pattern-match entities) export as name-value lists. Intercom supports custom attributes on contacts but does not have a native entity extraction feature equivalent to Certainly's NLU slot filling. We map entity names to Intercom contact custom attributes and flag any regex or pattern-based entities needing manual recreation as Fin AI topics or custom article sections.

Static response texts, rich message blocks, and button templates export cleanly from Certainly as structured JSON. We map them to Intercom's Message Templates and saved replies, preserving formatting, variable placeholders, and conditional content blocks. Templates that reference Zendesk ticket fields require substitution with Intercom conversation variables during migration.

Certainly's Zendesk connector configuration (ticket creation triggers, field mappings, agent assignment rules) is platform-specific and does not transfer to Intercom. We document the existing Zendesk integration settings during discovery as a written inventory so your admin can evaluate whether to recreate equivalent routing in Intercom's workflow builder or maintain Zendesk as a separate ticketing layer. This documentation is delivered as part of the migration scope.

Historical chat transcripts export from Certainly with session IDs, timestamps, agent attribution, and resolution status. Intercom's conversation export schema does not align with Certainly's log structure, and Intercom's reporting engine does not ingest historical conversation data from external sources. We archive logs as structured JSON alongside the migration, preserving all metadata for compliance and audit purposes. Spot-validation of archived records against source is performed before cutover.

Bot performance metrics, intent accuracy scores, conversation summaries, and deflection rate reports export from Certainly as flat files. Intercom's analytics are built from its own conversation data and cannot be seeded with external performance history. We archive these files in the customer's chosen storage format so historical performance data remains accessible for business intelligence tools.

Certainly operator and team assignments in routing rules map to Intercom Teammates. We resolve Certainly agent emails against Intercom workspace members, flagging any roles without a direct Intercom equivalent (such as Zendesk-specific escalation tiers). Role-based routing rules requiring Zendesk group logic are documented for manual recreation in Intercom's team and inbox settings.

Certainly's bot-level settings including greeting messages, session timeouts, proactive trigger rules, and messenger styling export as configuration data. We map these to Intercom's Messenger settings and bot configuration panel. Styling attributes (colors, logos, launcher text) transfer directly; proactive message triggers require manual recreation in Intercom's workflow builder because Intercom uses event-based rules rather than Certainly's time-based proactive triggers.

Certainly fallback policies define what happens when intent classification confidence is low or a user request falls outside trained flows. Intercom Fin AI Agent handles fallback by routing to a human inbox or triggering a resolution step. We document the existing fallback policy hierarchy so your admin can configure equivalent fallback routing in Intercom. We do not automate this rebuild as it requires understanding of your desired escalation path.