The product that made your data warehouse useful. Finally.
Hightouch invented Reverse ETL, built the leading Composable CDP, and is now an agentic AI marketing platform. Three ex-Segment engineers, $100M ARR, $2.75B valuation. The quiet giant of the modern data stack.
In April 2026, Hightouch announced that it had crossed $100 million in annual recurring revenue. At the same time, it closed a $150 million Series D at a $2.75 billion valuation. The investors - Goldman Sachs Asset Management and Bain Capital Ventures - were not betting on a startup. They were betting on a category.
Hightouch is what happens when you give marketers direct access to their company's data warehouse without requiring them to open a Jira ticket first. The platform connects sources like Snowflake, Databricks, and BigQuery to 250+ business destinations - Salesforce, Braze, Google Ads, HubSpot - and lets teams sync, segment, and activate that data in real time. No separate data store. No duplication. No data engineering backlog standing between a campaign idea and a campaign launch.
That's the product. The company that built it is stranger and more interesting than the pitch deck suggests.
"We coined the term Reverse ETL. We built the category. And now we're building what comes after it."
- Tejas Manohar, Co-CEO, HightouchBy 2018, most mid-to-large companies had a data warehouse. They had Snowflake or BigQuery or Redshift humming along, storing every transaction, every click, every user event. They had clean, reliable, well-governed data. And none of it was doing anything.
Meanwhile, the marketing team was running campaigns on stale CSV exports. Sales reps were looking at Salesforce records with three-day-old data. Ad teams were targeting audiences that were weeks out of date. The warehouse had everything. The tools that mattered had nothing.
The existing solution was a CDP - a Customer Data Platform. Platforms like Segment would collect your data, unify it, and push it to downstream tools. The problem: a CDP required you to move all your data into a new system. A separate store, a separate schema, a new vendor to trust with your most sensitive customer information. For companies that had already invested heavily in building out a warehouse, this felt less like a solution and more like a shakedown.
Three engineers who had worked at Segment - the company that arguably created the CDP category - looked at this situation and came to an obvious conclusion: the whole thing was backward. If you already have a warehouse with good data, you don't need to ETL it into a CDP. You need to do the reverse. You need Reverse ETL.
They named the category, built the product, and turned a niche data engineering workflow into a $2.75 billion company.
"Marketers are drowning in data they can't use. Hightouch is the pipe that makes it usable."
- Kashish Gupta, Co-CEO, HightouchTejas Manohar moved to San Francisco at 16. He was one of the first 10 engineers at Segment - meaning he watched from the inside as the CDP category was built, scaled, and eventually sold to Twilio for $3.2 billion. He left knowing exactly what was wrong with the model.
Josh Curl came from Michigan State with a CS degree and co-founded Deviceplane, an IoT fleet management company, before joining Segment. He became Hightouch's CTO and the engineering mind behind the warehouse-native architecture.
Kashish Gupta took the unusual path: Wharton MBA, venture capital at Bessemer Partners, then a sharp turn into Segment engineering. He and Manohar now run the company as co-CEOs - a structure that's unusual enough to invite skepticism and apparently works well enough to keep Goldman Sachs writing checks.
All three met at Segment. All three left to build something Segment couldn't - or wouldn't. They went through Y Combinator in 2019 and have not looked back.
One of the first 10 engineers at Segment. Moved to SF at 16. Co-coined "Reverse ETL." Now runs a $2.75B company at an age when most people are still doing performance reviews.
Michigan State CS graduate, ex-Segment, ex-Deviceplane co-founder. Designed the warehouse-native architecture that lets Hightouch activate data without ever touching a copy of it.
Wharton MBA, ex-Bessemer Ventures, ex-Segment. His unusual VC-to-engineer-to-CEO trajectory makes him the person Bain Capital trusted enough to back through four funding rounds.
Hightouch's product line has grown from a single Reverse ETL connector into a full suite of data activation capabilities. The common thread is the same as it was in 2019: all of it runs directly on your data warehouse, without ever moving data somewhere it doesn't belong.
SQL-based and no-code syncs from any warehouse to 250+ tools. The original product. Still the foundation. Syncs run incrementally so you're always working with fresh data.
Unified customer profiles, identity graphs, and audience segments built directly on your warehouse. All the capabilities of a traditional CDP - none of the data migration headaches.
Agentic AI that runs live experiments across your entire customer base, automatically choosing the best message, channel, and timing for each individual. Launched August 2024.
A no-code audience builder for marketers. Define, preview, and sync audience segments to any destination without writing SQL or waiting on data engineering.
Central coordination layer for multi-channel campaigns. Manage audience overlaps, suppression logic, and cross-channel orchestration from a single interface.
Server-side and client-side event collection that writes directly to your warehouse. Real-time behavioral data without a separate event pipeline or vendor relationship.
Total raised: ~$402M across 7 rounds. Series D was the largest single round at $150M, signaling institutional confidence in the AI marketing platform transition.
| Round | Amount | Date | Lead Investors |
|---|---|---|---|
| Seed / YC | ~$500K | 2019 | Y Combinator |
| Series A | $12.1M | Jul 2021 | Amplify Partners, Y Combinator |
| Series B | $40M | Nov 2021 | ICONIQ Growth |
| Series B Ext. | $38M | Jul 2023 | Bain Capital Ventures |
| Series C | $80M | Feb 2025 | Sapphire Ventures, Salesforce Ventures |
| C Extension | undisclosed | Jul 2025 | Snowflake Ventures, Capital One Ventures |
| Series D | $150M | Apr 2026 | Goldman Sachs Asset Management, Bain Capital Ventures |
Hightouch's customer list reads like a cross-section of modern enterprise: media companies, financial services firms, retailers, and SaaS businesses. The common denominator is that they all had data warehouses and found that getting data out was harder than getting it in.
Warner Music Group uses Hightouch to sync artist fan data to marketing tools. Chime uses it for financial product personalization. PetSmart runs audience targeting campaigns through it. Spotify, GameStop, Grammarly, Autotrader, Domino's, Monday.com - more than 30 documented case studies are publicly available.
"Philip Sonneveldt at Otrium went from a 4-week campaign cycle to 1 week - generating hundreds of ad concepts in a single session and driving a 15% lift in conversions with Hightouch Ad Studio."
- Hightouch Customer Case Study, 2026One of the more telling facts about Hightouch is that Snowflake Ventures invested in it - and then Snowflake named it Partner of the Year. Databricks awarded it ISV Partner of the Year in retail. Google Cloud co-published case studies about its Composable CDP. Capital One Ventures, a bank's strategic investment arm, put money in and became a customer.
This is not typical vendor-ecosystem backslapping. These are the companies whose platforms Hightouch runs on top of - and they're actively directing their customers toward Hightouch because Hightouch makes their platforms more valuable. Fivetran brought data in; Hightouch sends it out. dbt transforms data; Hightouch moves it downstream. The modern data stack is modular by design, and Hightouch is the module that closes the loop.
The stated mission is to make every company's data useful - to activate it directly from the warehouse into the tools teams already use, without copying, duplicating, or locking data into a separate platform. That's a clean sentence for a slide deck. The actual argument is more interesting.
Data warehouses were built by data engineering teams to serve analytics use cases. For years, that was their job. You queried them, you built dashboards, you ran reports. Then companies like Segment arrived and said: why not also use that data to personalize a customer experience, or trigger a marketing campaign, or suppress an ad for a customer who already bought the product? The answer was: because the warehouse wasn't designed for that.
Hightouch's bet is that the warehouse should be the system of record for all customer data - not just analytics - and that the right architectural pattern is to activate it directly from there. Not because it's philosophically elegant, but because companies have already spent millions building their warehouses and don't want to duplicate everything into a third-party system.
With AI Decisioning, the argument gets sharper. Instead of a data engineer writing a query and a marketer building a segment and someone scheduling a campaign - all of which takes days - Hightouch can watch behavioral signals in real time and make personalization decisions at the individual level, continuously, without human intervention. The marketer sets the guardrails. The AI runs the experiments. The warehouse provides the facts.
"The modern data stack was built to get data in. We built the other half - getting it out and into action."
- Josh Curl, CTO, HightouchIn April 2026, the same month Hightouch crossed $100M ARR and closed its Series D, it was also named a Leader in the Gartner Magic Quadrant for CDPs. The report - an industry benchmark that enterprise buyers use to vet vendors - placed Hightouch alongside the large marketing clouds it was built to compete against.
The company that started as a niche engineering tool for syncing Salesforce from Snowflake is now in the same evaluation framework as Adobe, Salesforce, and Microsoft. That's the proof of concept for a category that didn't exist in 2019.
What began with three engineers who were frustrated by the gap between where data lived and where it needed to go has become the infrastructure that Warner Music Group uses to communicate with fans, that Chime uses to personalize financial products, and that hundreds of other companies use to close the loop between data and action.
The warehouse is still there. The data is still in it. The difference is what happens next.