REPORT
The Modern Marketing Data Stack 5th Edition
Governing the Agentic Enterprise
The fifth edition follows the evolution of the modern marketing data stack:
- AI agents move from experiment to execution, with governance a key factor for scaling
- Organizations reorganize their stacks around governed data over applications
- Composability, trust and control are architectural priorities
From tools that assist to agents that act
AI has moved beyond features and into workflows. Systems now select tools, coordinate actions and shape decisions with limited human intervention. That changes the operating model for every marketing organization.
How do you lock down customer data for privacy and compliance while using it everywhere it creates value?
Who governs decisions when the decision-maker isn't human?
How do you build AI systems that are effective, accountable and trusted — not just fast?
The report examines how marketing stacks are being redesigned to answer these questions, and what separates organizations making real progress from those still experimenting.
At enterprise scale, AI makes one shift unavoidable: moving from fragmented, application‑led stacks to systems designed around outcomes. The stack must be governed as an operating model — rewiring data, processes and decisions to continuously reinvent how the business operates."
Elise Cornille
The stack reorganizes around the data
The systems connecting tools are becoming as important as the individual applications themselves. As AI-driven workflows span analytics, activation, measurement and collaboration, the defining question has shifted from "Which applications are in the stack?" to "Can the stack support coordinated decisioning, governed data access and controlled automation?"
The result is a structural shift:
Composable architectures are replacing or joining monolithic designs, but composability without governance creates its own risks
AI has created a new control plane above existing tools: a layer that assembles context, guides decisions and drives action across the stack
Shared semantics, identity resolution and consent enforcement are increasingly becoming foundational capabilities for organizations deploying AI at scale
Learn the architectural patterns emerging across successful deployments, the categories and vendors shaping the modern stack, and the governance foundations required to make it all work.
Operating the agentic enterprise in an uneven reality
Not every organization is ready for agents — and that's not a failure. Maturity varies by industry, scale, regulatory exposure and data readiness. The organizations making durable progress have something in common: they’re the ones that got governance, semantics and trust right first.
In this report, you'll learn:
Why moving too fast toward automation without data control introduces risk, not efficiency
Where hybrid states (legacy tools alongside composable components) are the norm, not the exception
How AI is changing collaboration — bringing cross-functional expertise into the work earlier
This report maps the patterns that work across different levels of maturity, with recommendations for wherever your organization sits today.
What I didn't fully anticipate was how much it would accelerate collaboration. A writer can rough out a storyboard in minutes and have a sharper conversation with an art director. Ideas move faster."
Scott Mager
Privacy, consent and control in the age of agents
Consumer privacy concerns aren't going away. Regulations are multiplying. And now AI agents are making decisions on customer data at a speed and scale that manual governance can't keep up with.
Universal opt-out signals are becoming law. Automated deletion requirements are expanding. Organizations are embedding compliance controls directly into infrastructure rather than treating them as downstream processes.
AI introduces new exposure: unintended data retention, unauthorized secondary use, models trained on data they shouldn't touch.
The organizations getting this right treat privacy as an operational capability — embedded in how data is accessed, shared and acted upon.
Explore how privacy, consent and governance are converging into a single architectural challenge — and what leading organizations are doing about it.
Without trustworthy, well-governed data, every AI ambition becomes noise and is slow, risky and impossible to scale. The teams winning today are the ones modernizing data foundations before anything else."
Kate Mackie
The Modern Marketing Data Stack 5th Edition
To discover leading solutions across the 13 categories that comprise the marketing data stack — and understand emerging martech trends — download the full report today.
Download the report
Methodology
Study objective and scope
The objective of this research was to identify the marketing technologies that have established a substantial and active customer base within Snowflake.
The study analyzed usage data and related trends over a 12-month period aligned to Snowflake fiscal year 2026, from Feb. 1, 2025, to Jan. 31, 2026. The analysis included data from more than 11,500 Snowflake customers.
Customer population included in the analysis
The analysis assessed usage by Snowflake active customers during the study period.
For the purposes of this research, active customers were defined as customers who held either a capacity or an on-demand/self-service contract with a valid end date, and generated revenue for Snowflake within the prior year.
All references to customer counts, usage, growth and rankings are based exclusively on this defined population.
Technology eligibility and inclusion criteria
Only technologies meeting the following eligibility requirements were considered for evaluation:
Active members of the Snowflake Partner Network (SPN), or
Operating under a comparable agreement with Snowflake, or
Snowflake Marketplace providers that have agreed to the applicable Marketplace terms and conditions.
Special category consideration:
Companies referenced in the large language models (LLMs) category represent providers whose LLMs are available in Snowflake's AI Data Cloud through Snowflake Cortex AI. These companies were not evaluated as part of the modern marketing data stack.
Categorizing technologies by type of Snowflake consumption
Vendors integrate with Snowflake's AI Data Cloud in different ways depending on their product architecture and use cases. To assess adoption and market leadership, technologies were categorized based on how they consume Snowflake.
The analysis evaluated technologies across the following consumption categories:
Technologies using Snowflake for data integration, transformation and analysis workloads
Technologies using Snowflake's collaboration workloads to integrate their product offerings, excluding data collaboration solutions
Technologies using Snowflake's collaboration workloads to provide a data collaboration offering, which corresponds to the collaboration category in this report
Each category uses distinct metrics and weighting to reflect the nature of Snowflake usage associated with that integration model.
Metrics used to measure adoption and success
Adoption success was measured using category-specific metrics aligned to how partner technologies leverage Snowflake capabilities.
Metrics for data integration, transformation and analysis workloads
For technologies using Snowflake primarily for data integration, transformation and analysis, the analysis measured:
Total number of active customers using the technology
Total Snowflake credit consumption attributable to the technology
Percentage growth in both customer count and credit consumption
Metrics for product integrations (excluding data collaboration solutions)
For technologies using Snowflake's collaboration workloads to integrate product offerings (excluding data collaboration solutions), the analysis measured:
Total number of stable edges that include the technology
Total Snowflake credit consumption driven by data shares used by the technology
Growth in both stable edges and credit consumption
A stable edge represents an ongoing relationship between a data provider and a data consumer. A stable edge is defined as a data share that:
Produced at least 20 transactions in which compute resources were consumed, and
Generated recognized product revenue,
Across two successive three-week periods, with at least 20 transactions in each period.
Metrics for data collaboration offerings
For technologies providing data collaboration offerings, the analysis measured:
Total number of stable edges that include the technology
Total Snowflake credit consumption driven by data shares used by the technology
Total number of unique consumers, defined as consumers with at least one stable edge established with a data provider
Growth calculation methodology
Growth percentages reflect changes in usage for each category's technologies over the 12-month trailing period from Feb. 1, 2025, to Jan. 31, 2026, compared with the immediately preceding 12-month period.
The growth analysis included all active capacity and self-service customers as of Feb. 1, 2025, who generated positive revenue for Snowflake during the subsequent year.
Construction of the Snowflake penetration index
To assess both market reach and usage intensity, Snowflake developed a penetration index that reflects how deeply and broadly technologies interact with Snowflake.
Index calculations apply weighted criteria tailored to each category of Snowflake consumption.
Breadth and depth are calculated during the 12-month analysis period.
1. Data integration, transformation and analysis index
This index combines breadth, depth and growth using the following weights:
Breadth (40%): Number of active customers using the technology on Snowflake
Depth (40%): Total Snowflake credit consumption attributable to the technology
Growth (20%), comprised of:
10%: Percentage growth in active customers
10%: Percentage growth in Snowflake credit consumption
2. Product integration (non-data collaboration) index
This index applies the following weights:
Breadth (40%): Number of stable edges associated with the technology
Depth (40%): Total Snowflake credit consumption attributable to the technology
Growth (20%), comprised of:
10%: Percentage growth in stable edges
10%: Percentage growth in Snowflake credit consumption
3. Data collaboration offering index
This index emphasizes market reach while accounting for usage depth:
Breadth (70%), comprised of:
40%: Number of stable edges associated with the technology
30%: Number of consumers integrated through the vendor's data collaboration offering
Depth (30%): Total Snowflake credit consumption attributable to the technology
Ranking and selection of marketing technologies
Technologies were ranked from 1 to N, where a lower rank indicates stronger performance.
Snowflake combined rankings across applicable metrics and normalized results on a 0-100 scale, where a score of 100 represents a technology ranked first across all applicable metrics.
Leader designation
Technologies were identified as Leaders based on the following criteria:
All categories except collaboration:
Ranked within the top three in their category, or
Achieved a technology index score within 87% of the average index score of the top three technologies in that category
Collaboration category:
Ranked within the top three in the category, or
Achieved a technology index score within 96% of the average index score of the top three technologies in the category
Ones to Watch designation
Ones to Watch were selected based on strong relative standing in the index, combined with qualitative factors such as:
Strong recent momentum in the market
Innovative technology or integration approach with Snowflake
Demonstrated customer capabilities
This methodology is intended to provide a transparent, consistent and defensible view of the marketing technologies gaining adoption and demonstrating impact within Snowflake's customer ecosystem.