Reverse ETL and a composable CDP that queries your data warehouse and writes audiences, fit scores and product usage into Salesforce, HubSpot, ad platforms and around three hundred other tools.
What is
Hightouch
Hightouch is a reverse ETL tool, which means it issues a query against the data warehouse your company already pays for and writes the results into the tools people work in. Salesforce, HubSpot, Outreach, Google Ads and around three hundred other destinations are on the other end of that sync. On top of that base are the pieces that make it a composable CDP. Customer Studio builds audiences with filters instead of SQL, Identity Resolution stitches profiles together, Events collects site behaviour, and Match Booster lifts ad match rates. Hightouch keeps no copy of the data, reading the warehouse every time. The newer bet is AI Decisioning, where a marketer sets goals, guardrails and content, and an agent picks the message, channel and send time for each customer, then keeps learning from what comes back. Gartner named Hightouch a Leader in its Customer Data Platforms Magic Quadrant in January 2026. Hightouch earns its price once your customer data lives in Snowflake, BigQuery or Databricks and someone in house writes SQL. It does from the warehouse what Clay does from a table, and at enterprise row counts that is the sturdier place for it. Most clients we build for have no warehouse, so their scoring goes through Clay.
Integrations and the modern GTM stack
Hightouch reads from a short list of sources and writes to a long list of destinations, so the integration story splits cleanly in two. Sources are warehouses and databases, and the roughly three hundred destinations are where the GTM value shows up.
Snowflake, BigQuery, Databricks and Redshift | The source side also covers Azure Synapse, Postgres, MySQL, SQL Server, MongoDB, S3, SFTP and Google Sheets. A model is a SQL query you write, a table you select, or a dbt, Looker or Sigma model that Hightouch imports and keeps in step. |
Salesforce | Standard and custom objects across Sales, Service, Health and Financial Cloud, in insert, update, upsert or archive mode. It uses the original Bulk API by default with Bulk API 2.0 as an opt-in, and the throughput ceiling is your Salesforce API allocation rather than anything Hightouch controls. |
HubSpot | Contacts, companies, deals and custom objects with insert, update and upsert, plus custom behavioural events on insert and contact lists on add or remove. Association mapping is configured in the sync. Two syncs at once will trip HubSpot's per-second rate limit, so stagger the schedules. |
Outreach and Salesloft | Outreach takes accounts, prospects and sequence states, so a warehouse model can enrol somebody in a sequence by combining prospect, sequence and mailbox IDs. Accounts and prospects support upsert while sequence states are insert only, which is the piece that turns a fit score into activity. |
Meta, Google, LinkedIn and TikTok | Audiences built in Customer Studio sync straight to the ad platforms, and Match Booster appends personal email, work email, phone, mobile IDs and address before the handoff. Hightouch claims match rate lifts of 48 percent on Facebook, 35 percent on TikTok and 24 percent on LinkedIn. |
dbt, Looker, Sigma and Fivetran | These are in the extensions tier rather than the destination catalogue. They let you pick models from a git repo or a BI tool as the input to a sync, and trigger the sync as soon as a dbt Cloud or Fivetran job finishes so the CRM never reads a half-built table. |
Slack, Zapier and HTTP Request | Slack carries sync alerts and record-level notifications for a rep channel. The HTTP Request destination posts rows to any API you can authenticate against, and the Zapier route goes through a webhook trigger with Zapier's own ceiling of 10,000 requests every five minutes. |
Clay | There is no Clay connector in the catalogue, and hightouch.com/integrations/destinations/clay returns a 404. The working route is the HTTP Request destination posting rows into a Clay table through Clay's API, which functions fine and needs somebody to own the payload and the match key. |
What is missing | No Instantly, Smartlead or Lemlist destinations, so cold email infrastructure gets fed through the webhook destination or through the CRM record. Anything below hourly latency is also on a quoted tier, since the free and self-serve plans cap sync frequency at once an hour. |
Pro’s
The warehouse stays the source of truth because Hightouch stores no copy of your data, so the definition of a qualified account lives in one modelled table instead of drifting across five tools.
Around three hundred destinations covers the CRM, the sequencer and the ad platforms together, so one model can feed Salesforce, Outreach and Meta without a separate connector project for each of them.
Customer Studio lets a marketer assemble an audience with filters and traits, which keeps the data team out of the loop for routine segment changes that used to arrive as tickets.
The free tier gives two active syncs with unlimited destinations and seats, so a team can prove the pattern on real data before anyone books a call with sales.
The dbt, Looker, Sigma and Fivetran extensions let a sync start the moment the model behind it finishes building, and that stops the tools downstream from reading yesterday's numbers.
Security coverage extends to SOC 2 Type 2, ISO 27001, HIPAA and GDPR with EU region deployment, so a regulated buyer has fewer reasons to stall the review.
Cons
Hightouch does nothing without a working data warehouse, so a team with no Snowflake, BigQuery or Databricks is looking at a data engineering project measured in months before the first sync happens.
Reverse ETL models are SQL, so the people who own the definitions are in the data team while the GTM operators who feel the pain wait in a queue behind them.
Hightouch publishes no dollar prices at all, and Vendr puts the median contract at 15,000 dollars a year across 176 purchases, which makes the real cost a negotiation rather than a number you can plan around.
The free and self-serve plans cap sync frequency at hourly, so any signal that needs to reach a rep within minutes lands you on a quoted tier straight away.
Reviewers on G2 keep flagging thin error messages and no way to describe a sync, so a quiet failure upstream can leave your CRM working from stale data with nobody watching.
Pricing
Basic Reverse ETL
$0/month Free tier, self-serve signup, no demo
Up to 2 active syncs
Unlimited destination count
Unlimited user seats
Hourly sync frequency at most
One workspace per organisation
Self-serve
Quoted No dollar figure published by Hightouch
10 active syncs
No limit on destinations, seats or sync runs
Hourly sync frequency at most
100,000,000 operations a month cap
One workspace per organisation
Composable CDP
Usage based, demo required Pick the products you want, priced on usage
Reverse ETL, Events, Customer Studio
Identity Resolution, Match Booster, Real-time Personalization
Unlimited user seats and premium extensions
Advanced observability and access management
Dedicated support and custom SLAs
Agentic Marketing Platform
Usage based, demo required The AI Decisioning bundle for lifecycle and ads
Ad Studio and Lifecycle Studio
AI Decisioning and Content Assembly
Pre-set amount of monthly personalised actions
Integrations with any marketing platform
Enterprise security and observability
How it fits your stack
Alternatives
At a glance
Reverse ETL from Snowflake, BigQuery, Databricks
Around 300 destinations including Salesforce and HubSpot
Customer Studio audience builder needs no SQL
Identity Resolution, Events and Match Booster
AI Decisioning picks message, channel and timing
dbt, Looker, Sigma and Fivetran extensions
Free tier with two active syncs
SOC 2 Type 2, ISO 27001, HIPAA, GDPR
Read some of our client stories
GTM Workflows
Learn how to use GTM tools in workflows
Guides and tutorials to get the most from your workflows
Ready to get started?
By clicking Sign Up you're confirming that you agree with our Terms and Conditions.
