Customer.io Integrations: Native, Zapier, and API

The official directory listed 109 integrations as of August 19, 2026. Data In means sources that write into Customer.io: warehouses, databases, CDPs, form tools, reverse-ETL partners and the mobile SDKs. Data Out means destinations Customer.io writes to: ad platforms, analytics, support desks, storage buckets and other messaging platforms. Segment, Salesforce, HubSpot, Snowflake and BigQuery appear on both sides, and run bidirectionally.

Journeys is the messaging product: profiles, automations, one-time sends, transactional messages and the Deliveries view. Data Pipelines is the CDP layer that collects, cleans and fans data out. They use different APIs and different credentials. Journeys ingestion is the Track API at track.customer.io with HTTP Basic (Site ID as username, Track API key as password). Data Pipelines is the Segment-spec API at cdp.customer.io with HTTP Basic using the source write key as username and a blank password. Sending and management use a third API, the App API at api.customer.io with a Bearer App API key. The wrong key against the wrong host returns 401 with no hint about which key was expected.

As a source, add Customer.io as a Segment destination so identify and track calls land as profiles and events; Segment is also a Data In connection on the Pipelines side. As a destination, Customer.io writes profiles and message events back to Segment. Data Pipelines is itself Segment-spec compatible (identify, track, page, screen, group, alias, batch, all POST-only), so an existing analytics library can often be repointed at cdp.customer.io instead. Batch payloads cap at 500 KB total with 32 KB per contained call.

Yes, both are bidirectional: CRM records sync in as profiles, and audiences plus messaging activity sync back onto CRM objects. Check the billing model before you switch it on. Customer.io meters profiles, defined as people plus objects, and objects include accounts, companies and devices. A B2B sync of 5,000 contacts plus 3,000 accounts consumes 8,000 profiles against an Essentials plan that starts at $100 per month for 5k profiles. Size the object model before the first full sync, not after.

Each has a dedicated Data Out destination — Google Ads Conversions, Meta Conversions API and TikTok Conversions, with Pinterest and Bing Ads also available. All are server-side, so conversions and audience membership leave Customer.io directly rather than through a browser pixel. Two things need care: the identity field you match on (hashed email is the usual choice) and the event mapping, since a Customer.io event name will not equal an ad-platform conversion action by default. If a platform is missing, the Custom Webhook destination can post to its conversions endpoint.

Three routes. The Reporting Webhook streams the full message lifecycle — sent, delivered, opened, clicked, bounced, spammed, unsubscribed, failed — to any HTTPS endpoint you control. The Snowflake, BigQuery, Redshift and S3 destinations export events and metrics on a schedule, which suits analytics better. The Slack destination posts on journey criteria, so it fits alerting. One limitation to plan around: if you enable custom SMTP so Customer.io relays through your own provider, Customer.io loses its feedback loop and bounce and delivery data stops appearing — unless your relay is Mailgun, Mailjet, Mandrill, Postmark, SendGrid or SparkPost with feedback configured.