Hightouch closed $150 million at a $2.75 billion valuation last week to build "agentic marketing" — and the round is a direct shot at Salesforce, Adobe, and every suite-embedded marketing cloud charging six-figure data tax to enterprises. Goldman Sachs Alternatives and Bain Capital Ventures led the Series D, with Iconiq Capital, Sapphire Ventures, Amplify Partners, Y Combinator, and The Trade Desk's TD7 fund piling on. The valuation jumped 2.3x from the company's $1.2 billion Series C just 14 months ago, after Hightouch posted 100%+ revenue growth in each of the past two years.
For CIOs, CMOs, and CFOs evaluating their 2026 marketing technology stacks, this round signals more than a single startup's success. It signals that the entire architecture of enterprise marketing is being rebuilt around the data warehouse — not the suite — and that AI agents are the layer that finally makes "composable CDP" a buying decision instead of an engineering project. The buying question for 2026 is no longer "which marketing cloud" but "do we deploy agents on data we already own, or pay $360,000+ per year to copy that data into someone else's vendor stack?"
What Changed: The Deal and Why It Matters Right Now
The April 29 funding announcement combined three signals that rarely line up at once: a category-defining valuation jump, blue-chip enterprise customer wins, and a clear competitive thesis against the largest marketing cloud incumbents.
Deal mechanics. Hightouch raised $150 million in Series D financing, co-led by Growth Equity at Goldman Sachs Alternatives and Bain Capital Ventures, with continued participation from Iconiq Capital, Sapphire Ventures, Amplify Partners, Y Combinator, and The Trade Desk's venture arm TD7. The $2.75 billion post-money valuation more than doubles the $1.2 billion mark set in the company's $80 million Series C in February 2025 (source).
Customer roster. The customer logos disclosed in the round include Domino's, PetSmart, DraftKings, Ramp, WHOOP, Autotrader, Cars.com, Aritzia, and Spotify (source). These are not pilot accounts. They are public reference customers operating across retail, food service, sports betting, fintech, automotive, and DTC commerce — segments where Salesforce Marketing Cloud and Adobe Experience Cloud have been the default choices for over a decade.
Investor thesis. Goldman Sachs partner Darren Cohen put the rationale plainly: "AI is fundamentally changing how enterprises operate, and marketing is one of the largest functions poised for transformation" (source). Goldman explicitly framed Hightouch's bet on deploying AI agents directly on customer data warehouses as the architecture that "defines the next category of marketing infrastructure."
Co-CEO framing. Kashish Gupta, Hightouch's co-founder and co-CEO, was even more direct in dismissing the current state of marketing AI: "Marketing is sorely in need of reinvention, but most AI solutions haven't actually changed how marketing works" (source). Co-founder and co-CEO Tejas Manohar described the agent experience as "going into a bank and meeting a teller who knows you. They're not just going to minute-one tell you about offer X and keep repeating it" (source).
Why now. Hightouch is shipping six core products: Ad Studio (creative generation at scale), Content Assembly (asset remixing), a Composable CDP (warehouse-native customer data), Reverse ETL (sync to downstream channels), Real-time Personalization, and an Intelligence Platform for analytics (source). The agentic layer sits across these, observing signals and executing the next-best-action within enterprise-defined budget caps, brand-safety rules, and approval workflows. That mix is what separates "AI inside marketing software" (every vendor has it) from "agents that actually run campaigns end-to-end" (very few have shipped it).
The funding announcement landed in the same week Anthropic and OpenAI both raised multi-billion-dollar rounds to build forward-deployed AI services firms (Anthropic vs OpenAI services war). The pattern across these moves is consistent: in 2026, enterprise AI value capture is migrating from horizontal model APIs to vertical, deeply-integrated workflow platforms — and marketing is the largest functional budget being repointed.
Why This Matters: The Technical and Business Implications
This is a story with two distinct readers — and the buying decision lives at the intersection of both.
Technical Implications (CIO, CTO, Chief Data Officer)
Data gravity wins. Suite-embedded CDPs like Salesforce Data Cloud and Adobe Real-Time CDP require enterprises to copy customer data out of their authoritative source — typically Snowflake, Databricks, BigQuery, or Redshift — into the vendor's proprietary platform. Composable, warehouse-native CDPs like Hightouch invert this: data stays in the warehouse, and activation happens via reverse-ETL to downstream tools. For data engineering teams, this eliminates the "two sources of truth" problem and dramatically simplifies governance.
Governance and lineage. With customer data living in the warehouse, existing access controls, masking policies, retention rules, and audit logs already cover marketing use cases. There is no separate identity stitching layer to harden, no parallel pipeline to monitor, and no second compliance scope for SOC 2, GDPR, or CCPA reporting. For regulated industries — financial services, healthcare, insurance — this is the difference between a six-month deployment and a six-week deployment.
Agent control plane. Hightouch's agentic layer is built around three primitives every enterprise IT leader should understand: trusted data context (the warehouse), business logic and constraints (budget caps, bid ranges, brand-safety rules), and the action surface (paid media, email, SMS, web personalization). This is the same agentic architecture pattern emerging across the enterprise AI stack — observe, plan, act, learn — but specialized for marketing workflows (source).
Integration burden. Suite-embedded CDPs assume the rest of your stack is also from the same vendor. If your CRM is Salesforce but your commerce is Shopify and your support is Zendesk, suite CDPs become integration tax centers. Warehouse-native CDPs treat every downstream system as a sync target, which lowers the integration risk for heterogeneous stacks (the reality at almost every Fortune 1000).
Business Implications (CFO, CMO, COO)
Marketing operations productivity is now measurable. Recent data shows marketing operations functions delivering 3.1x productivity gains from AI, with marketers saving an average of 11 hours per week and 76% of organizations achieving marketing automation success within one year (source). That moves AI investment out of "experimental R&D" and into "core operations" — the same reclassification JPMorgan made when it moved AI to core infrastructure (JPMorgan AI infrastructure).
The TCO story flips the spreadsheet. A typical enterprise Salesforce stack — Data 360, Marketing Cloud, Service Cloud, MuleSoft, Agentforce, plus systems integrator fees — runs $360,000 to $750,000+ in first-year costs, with Data Cloud charging $420 per 1,000 enterprise profiles per year and Agentforce 1 Editions starting at $550 per user per month (source). Adobe Real-Time CDP pricing is quote-based but tied to profile volume, with full deployments typically requiring AEP, RT-CDP, Journey Optimizer, and Customer Journey Analytics as separate licenses. Warehouse-native architectures keep customer data costs inside the existing data platform budget — avoiding the duplicate profile fees entirely.
Revenue impact, not just cost savings. A Hightouch financial-services customer reportedly generated $50M+ in incremental annual revenue from advertising, launched ad creative 80% faster with ~10% expanded reach, and replaced 60 manually-built customer journeys while improving onboarding outcomes by 30%+ (source). For a CMO, those are the numbers that justify a board conversation, not a procurement debate.
Strategic positioning for CFOs. With 88% of organizations planning budget increases specifically for agentic capabilities and 62% expecting ROI exceeding 100% on those investments (source), the question is not whether to fund agentic marketing — it's which architecture earns the multi-year commitment. CFOs who lock into suite contracts in 2026 may be paying for capacity their warehouse-native peers have already amortized.
Market Context: A $58B Category Being Restructured in Real Time
The customer data platform market sits between $4.58 billion (Mordor Intelligence) and $10.49 billion (Grand View Research) in 2026, depending on whether you count just dedicated CDP software or include implementation services and adjacent martech tools. Most credible sources triangulate to roughly $5–6 billion in 2026, with IDC pegging the market at $5.7 billion+ (source). Long-term forecasts project a 27.8% CAGR, reaching $58.41 billion by 2033 — a 10x expansion in less than a decade.
What's changing is not the category size. It's the architecture inside the category.
Gartner's adoption forecast. By 2028, 60% of brands will use agentic AI to deliver streamlined one-to-one interactions, up from less than 5% in 2025. By the end of 2026, 40% of enterprise applications will include task-specific AI agents — and marketing is one of the top three functions named (source). The 2026 Gartner CIO Survey reports 17% of organizations have deployed AI agents and 60%+ expect to within two years.
The failure rate is real. Gartner also projects 40%+ of agent projects will fail by 2027, and only 5% of marketing leaders not yet piloting AI agents report significant business gains. Forrester sees 2026 as the breakthrough year for multi-agent systems, with sales and marketing agents producing 2-3x improvements in pipeline velocity for organizations that get the data foundation right.
Competitive landscape. Hightouch's nearest architectural peers are Census, RudderStack, and Twilio Segment in the composable CDP space. The suite-embedded competitors — Salesforce Data Cloud + Agentforce, Adobe Real-Time CDP + AI Assistant, Oracle Unity, and Microsoft Customer Insights — own the installed base but face an architectural tax: they require data duplication, vendor lock-in, and full-suite licensing to unlock AI features. The composable challengers can sell into Snowflake, Databricks, and BigQuery customers without disrupting the data layer at all.
The talent and governance gap. Deloitte research shows agentic AI is scaling faster than the guardrails meant to govern it: only 21% of IT and business leaders report a mature governance model for agents (source). For marketing specifically, brand-safety violations, policy escalations, and compliance risks rise sharply when agents are deployed without budget caps, action approval flows, and continuous monitoring. The vendors that win 2026 will be the ones that ship governance as a default — not as a customization project.
Framework #1: Three-Year TCO and ROI Calculator — Warehouse-Native vs Suite-Embedded
The TCO conversation between warehouse-native and suite-embedded CDPs is where most marketing technology decisions actually get made. Below are three enterprise scenarios with concrete math. All numbers are list-price-anchored estimates derived from public pricing pages; negotiated discounts of 20-40% are standard for enterprise contracts.
Scenario A: Mid-Market Retailer — 2 Million Customer Profiles, 50 Marketing Users
Salesforce Data Cloud + Agentforce path (3-year TCO):
- Data Cloud profiles: 2,000 × $420 = $840,000/year × 3 = $2,520,000
- Marketing Cloud Engagement Pro (50 users): ~$3,750/month × 36 = $135,000
- Agentforce 1 (50 users): $550/user/month × 50 × 36 = $990,000
- Implementation and integration services: $300,000–$500,000 one-time
- Total 3-year TCO: $3.95M–$4.15M
Hightouch warehouse-native path (3-year TCO):
- Existing Snowflake/Databricks cost (already spent): $0 incremental
- Hightouch platform fee (estimated mid-market tier): $120,000–$240,000/year × 3 = $360,000–$720,000
- Implementation services (warehouse-native = lighter integration): $80,000–$150,000 one-time
- Total 3-year TCO: $440,000–$870,000
Net savings: $3.1M–$3.6M over 3 years. Break-even: under 4 months.
Scenario B: Enterprise Financial Services — 15 Million Customer Profiles, 200 Marketing/CX Users
Salesforce stack 3-year TCO:
- Data Cloud profiles: 15,000 × $420 = $6.3M/year × 3 = $18.9M
- Service + Marketing Cloud licensing: $2.5M–$3.5M
- Agentforce 1 (200 users): $550 × 200 × 36 = $3.96M
- MuleSoft + integration services: $1.5M–$2.5M
- Total 3-year TCO: $26.9M–$28.9M
Hightouch + warehouse path 3-year TCO:
- Existing data warehouse: amortized
- Hightouch enterprise platform: $400,000–$750,000/year × 3 = $1.2M–$2.25M
- Brand safety, governance, monitoring tooling: $300,000 one-time
- Existing channel tools (paid media, email): unchanged
- Total 3-year TCO: $1.5M–$2.55M
Net savings: $24M–$27M over 3 years. Break-even: under 2 months.
Scenario C: Global Consumer Brand — 80 Million Profiles, 500 Marketing/CX Users, 12 Countries
Salesforce stack 3-year TCO:
- Data Cloud profiles: 80,000 × $420 = $33.6M/year × 3 = $100.8M
- Marketing Cloud + Service Cloud: $10M–$14M
- Agentforce 1 (500 users): $9.9M
- Multi-region deployment + SI fees: $8M–$12M
- Total 3-year TCO: $128M–$146M
Hightouch + warehouse path 3-year TCO:
- Hightouch enterprise platform with multi-region: $3M–$5M/year × 3 = $9M–$15M
- Governance, observability, regional compliance tooling: $2M–$3M
- Total 3-year TCO: $11M–$18M
Net savings: $110M–$128M over 3 years.
The savings ratio holds across enterprise sizes because suite-embedded CDPs scale costs linearly with profile count, while warehouse-native architectures push the data cost curve to zero (the warehouse is already paid for) and charge for orchestration capacity instead. For organizations with 5M+ profiles, the math is rarely close.
Framework #2: Agentic Marketing Readiness Assessment — Score Your Org 1-25
Before any vendor decision, marketing and IT leaders should score their own organization across five dimensions. Each dimension scores 1-5 (1 = not ready, 5 = production-ready). Total: 25 points.
Dimension 1: Data Foundation (1-5)
- 5 — Single source of truth in cloud data warehouse, identity resolution working, refreshed in near real time
- 3 — Warehouse exists but identity stitching is incomplete or batch-only
- 1 — Customer data scattered across siloed systems, no unified profile
Dimension 2: Channel Activation Maturity (1-5)
- 5 — Reverse-ETL pipelines or composable activation already running to email, ads, SMS, and web
- 3 — Some channels integrated, but most still require manual list pulls
- 1 — Marketing channels run on disconnected tools with manual exports
Dimension 3: Governance and Brand Safety (1-5)
- 5 — Documented brand-safety rules, budget caps, approval workflows, and policy violation monitoring already in place
- 3 — Some guardrails exist but enforcement is manual or post-hoc
- 1 — No formal guardrails; agents would operate without explicit constraints
Dimension 4: Measurement and Attribution (1-5)
- 5 — Closed-loop measurement from spend to revenue with statistically valid lift testing
- 3 — Last-touch attribution working but multi-touch is incomplete
- 1 — Marketing performance is reported in disconnected dashboards with no consolidated KPI
Dimension 5: Organizational Change Readiness (1-5)
- 5 — Executive sponsor named, change management plan in place, marketing team trained on AI workflows
- 3 — Leadership supportive but operational team skeptical or untrained
- 1 — No executive sponsor, no change plan, no training
Scoring guide:
- 20–25: Production-ready. Move to vendor selection and pilot in 30 days.
- 15–19: Partially ready. Pilot with a single use case (e.g., paid social ad creative). Fix gaps in parallel.
- 10–14: Not ready. Fix data foundation and governance before any agentic deployment.
- Under 10: Architectural rebuild required. Agentic AI will fail until the foundation exists.
This is the same scoring discipline that separates the 5% of marketing leaders seeing significant gains from the 40% whose agent projects fail by 2027. Most enterprises score 12–17 on this scale today — meaning the agentic marketing buying decision in 2026 is at least 50% an internal readiness decision and only 50% a vendor decision.
Case Study: A Financial Services Customer's $50M Incremental Revenue
A Hightouch financial services customer disclosed in the funding announcement provides a useful reference point for what production agentic marketing looks like at scale.
Starting state. The customer operated a traditional martech stack with siloed data, manually-built customer journeys, and creative production cycles measured in weeks. Marketing operations spent the majority of their week building audience segments, configuring routing rules, and waiting on creative agencies to produce variants for paid social and display.
What changed. The team replaced 60 manually-built customer journeys with agentic workflows operating directly on the data warehouse. Ad Studio took over creative generation, producing branded variants at scale and launching them into paid channels with continuous performance learning. Content Assembly handled asset remixing across email and onsite personalization. Real-time orchestration replaced the weekly batch routing logic with continuous next-best-action decisioning.
Quantified outcomes.
- $50M+ in incremental annual revenue from advertising
- Ad creative generation and launch cycles 80% faster
- Reach expanded by approximately 10% (same budget, better targeting)
- Customer onboarding outperformed prior methods by 30%+
- 60 manual journeys collapsed into orchestrated agent workflows
Lessons learned. The customer's success rested on three pre-existing capabilities: a unified customer data warehouse, clearly-defined business constraints (budget ranges, brand rules, compliance requirements), and a measurement framework that already linked marketing spend to revenue. The agentic layer multiplied existing maturity — it did not create maturity from nothing.
This pattern matches Forrester's broader finding that marketing AI agents produce 2-3x pipeline velocity improvements for organizations with strong data foundations, and almost no measurable lift for organizations that try to deploy them on top of broken plumbing. The gap between the top quartile and bottom quartile of agentic marketing programs in 2026 is not vendor selection — it's data and governance maturity.
What to Do About It
For CIOs and Chief Data Officers. Run the readiness assessment in the next 30 days. If your organization scores 15+, begin a vendor evaluation that includes both warehouse-native (Hightouch, Census, RudderStack) and suite-embedded (Salesforce, Adobe) options — but force vendors to compete on three-year TCO at your specific profile volume, not on demo features. If your score is below 15, fix the data foundation before any agentic marketing pilot. The data layer is the single largest determinant of agentic ROI.
For CFOs. Demand the three-year TCO model from marketing technology proposals, with explicit profile-volume scaling. Suite-embedded CDP costs scale linearly with profile count; warehouse-native costs scale with orchestration capacity. At 5M+ profiles, the gap is rarely under $10M over three years. Tie any agentic AI budget approval to a specific business outcome metric — incremental revenue, cost-per-acquisition reduction, or productivity hours saved — not vendor feature counts. Track the AI investment as core operations, not innovation R&D, the same way enterprises like JPMorgan have already reclassified theirs.
For CMOs and marketing leaders. Pilot one use case end-to-end with full agentic orchestration before scaling — paid social creative is the most-cited starting point because it's measurable, contained, and produces fast feedback loops. Define brand-safety rules, budget caps, and approval thresholds in writing before the first agent runs. Train marketing operations teams on the new workflow (segment-building → goal-setting and constraint-setting); the role shift is real and uncomfortable, but unavoidable.
For boards. Ask management two questions on the next earnings prep cycle: "What percentage of marketing operations is automated by agents today, and what is the trajectory?" and "What is our three-year marketing technology TCO under the current architecture versus a warehouse-native architecture?" The companies that can answer both with specific numbers are the ones positioned to compound the productivity gains over the next decade.
Want to calculate your own AI ROI? Try our AI ROI Calculator — takes 60 seconds and shows projected savings, payback period, and 3-year ROI.
