CDP Market Landscape 2026

Methodology: How we count

Introduction

Most CDP market reports are built from vendor briefings and analyst surveys. This one is built from documentation. Over 2026, CdpHub read the public docs and integration catalogs of 44 customer data and engagement platforms and recorded every integration each one documents: 6,699 listings across 2,365 tools, each with its direction (into or out of the platform), the way it runs (client-side tag or SDK, server-side API or file, or warehouse), its category and a link to the page that describes it.

That gives a different view of the market than the one vendors present. Integration catalogs are a platform's honest statement of what it was built to do: an ingestion-first platform documents sources, an activation platform documents destinations, a composable platform documents warehouses. The three camps the February edition of this report described are still the right map. The data shows where the borders actually are, which platforms sit on them, and how much of the "integration breadth" claim survives contact with a tool-by-tool count.

1. The market in one table

44 platforms, ranked by documented integrations. "Tools" counts distinct partner products (a platform with separate audience and conversion listings for Meta Ads counts Meta once). "Out" means data leaves the platform; "In" means it arrives. "Warehouse" is the share of listings that read or write a warehouse, lake or bucket, or run as reverse ETL.

* ActionIQ's listings beyond its documented connectors come from its catalog and have no run type. SAP CDP documents its integrations on SAP's help portal, whose robots.txt doesn't let our reviewers read it, so only the 2 integrations described on other SAP pages are listed. Every other platform was rebuilt in 2026 from its own documentation and catalog.

Two readings of the table:

• Raw count rewards tag managers. Tealium leads with 886 listings because 594 of them run client-side, nearly all of them tags: it documents every vendor pixel it can deploy. Segment's 555 is the largest server-side catalog. Counting only outbound integrations that run from the platform's own servers, the order is Segment (359), Tealium (249), mParticle (237), Klaviyo (168) and RudderStack (149); count reverse ETL too and Hightouch (300) moves into second place.
• Direction is the clearest fingerprint. Salesforce Data 360 documents 156 inbound connectors and 11 outbound. Klaviyo documents 231 sources. Tealium, mParticle, Zeotap and Zeta are 80 to 93 percent outbound, and so are the reverse ETL platforms (DinMo 89 percent, Fivetran Activations 88, Hightouch 86). Treasure AI (formerly Treasure Data), NICE, Redpoint and the marketing-suite CDPs (Bloomreach, Lytics) sit in the middle. A platform's catalog tells you whether it expects to be your system of record or your last mile.

2. Three camps, measured

The February edition sorted the market into packaged, composable and suite CDPs by architecture. The warehouse share in the table above is the number that separates them.

Composable (warehouse-native). Fivetran Activations (formerly Census) and DinMo run all but one integration each through the customer's warehouse (the exceptions are a Datadog monitoring feed and Slack alerts), and Hightouch 95 percent of its integrations. Hightouch's other 15 include stream sources for its Hightouch Events product, message sends for its AI Decisioning agents, and dbt, Fivetran, Datadog, PagerDuty and Slack hooks. They own no profile store and their catalogs are mostly destinations (Hightouch: 300 out, 46 in; Fivetran Activations: 172 out, 23 in). RudderStack is the hybrid: 18 percent warehouse, almost a third client-side, and 139 tools in common with Segment, the second-largest overlap of any pair in the directory after Hightouch and Fivetran Activations (143).

Packaged (platform owns the profile). Segment, mParticle, Tealium, Klaviyo, Braze, CleverTap, Lytics and the marketing-suite CDPs document fewer than 15 percent of integrations as warehouse connections. Their catalogs are built around their own event collection (client-side share: Tealium 67 percent, BlueConic 32, RudderStack 31, Adobe 30, Segment 20) and server-side fan-out.

Suite (CDP inside a larger stack). Adobe Experience Platform (279 listings, 10 percent warehouse), Salesforce Data 360 (177, 20 percent, almost entirely inbound), Microsoft Dynamics 365 Customer Insights (67, 16 percent), SAP (SAP Engagement Cloud 63 and SAP CDP 2) and Oracle (40). Their catalogs are smaller than the packaged leaders' because the suite's own products fill many of the gaps. SAP CDP's count is low because its integration docs sit on a help portal our reviewers can't read.

The border is moving. Salesforce Data 360 (20 percent warehouse, with 35 different warehouse, lake and database tools, the most of any platform), Treasure AI (20 percent), Redpoint (40 percent) and Simon Data (28 percent) are packaged or suite platforms that now document warehouse read-and-write as a primary path. That is the composable argument being absorbed rather than won.

3. Five trends, with numbers

The warehouse is the shared endpoint

1,222 of 6,699 listings (18 percent) read or write a warehouse, lake or bucket. Snowflake is documented by 35 platforms, Amazon S3 by 33, BigQuery by 30, Redshift by 28, Google Cloud Storage by 27, Databricks by 26 and Azure Blob Storage by 25. Among the 20 largest catalogs, Snowflake is on all 20 and BigQuery on 19. Every camp now treats the warehouse as an integration target; the camps differ in whether it is also the profile store.

Advertising is still the biggest destination

Advertising is the largest category with 1,308 listings, ahead of Email & Marketing Automation (905) and Data Infrastructure (829). Meta is documented by 38 platforms, Google Ads by 35, TikTok by 29, LinkedIn by 28, Snapchat by 24 and Criteo by 21. Audience activation to paid media remains the first job most CDPs document, whatever their architecture.

The core is small and the tail is long

90 tools appear on ten or more platforms; 1,653 of the 2,365 tools (70 percent) appear on exactly one. The common core is warehouses, ad platforms, Salesforce, HubSpot, Shopify, Zendesk, SendGrid and Google Analytics. Beyond it, catalogs diverge fast: the two most similar catalogs, Hightouch and Fivetran Activations (both reverse ETL), share 43 percent of their tools, and Segment and RudderStack share 26 percent. "Integration breadth" is real as a differentiator, but only the long tail differentiates, so the test for a buyer is whether a platform documents the specific tools in their stack, not its headline count.

Catalogs reveal specialization

The leading category per platform is a better taxonomy than vendor positioning. NICE's catalog is Customer Support (52 of 165 listings). CleverTap's is Mobile & Push (52 of 172). 6sense, CaliberMind and Leadspace are CRM & Sales. Klaviyo, Listrak, Bloomreach, Optimizely, Ometria and SAP Engagement Cloud are E-commerce. Salesforce Data 360, Treasure AI, Hightouch, Fivetran Activations, Redpoint, Odie, Snowplow, Simon Data, Informatica and Oracle are Data Infrastructure, and Teradata's partners are mostly analytics and BI tools. As partners, Braze (listed by 23 of the other platforms) and Klaviyo (19) are the most connected platforms in the directory; Segment (17) is the most connected CDP, well ahead of mParticle and Tealium (9 each).

Consolidation shows up in the documentation first

Several catalogs changed hands or names while CdpHub was reading them. ActionIQ now documents under Uniphore and BlueVenn under Upland, and Blueshift's documentation says it is part of BlueConic. Census joined Fivetran as Fivetran Activations, Totango is now Odie, CrossEngage is now Spotler ActivatePro, and SAP renamed SAP Emarsys to SAP Engagement Cloud in February 2026; among the tools they connect to, Ninetailed is now Contentful Personalization, Selligent is Marigold Engage and Outbrain became Teads. The directory shows each under its new name, with the old one alongside. Delighted shut down on June 30, 2026, and Oracle closed its advertising business (BlueKai, Datalogix, Moat) in 2024. Their listings are gone, along with those for 86 other partner products that have shut down or been folded into an acquirer, such as AddThis, Bronto, Flurry and Polyvore.

4. How integrations run

Of 6,699 listings, 1,221 run client-side (a tag, pixel or SDK in the site or app), 4,231 server-side (API, webhook, stream or file transfer), 1,222 through a warehouse, and 274 have no documented run type. By tool: 746 tools have a client-side path on at least one platform, 1,634 a server-side path, 438 a warehouse path.

Client-side share is the tag-management inheritance: Tealium 67 percent, Chord 40, BlueConic 32, RudderStack 31, Adobe 30, Segment 20, mParticle 16. Platforms built after third-party cookies came under pressure document almost no client-side paths (zero for Fivetran Activations, Amperity, Salesforce Data 360 and NICE, and one for Hightouch). For a buyer, this is the privacy and performance question in one number: how much of a platform's catalog depends on code running in the browser.

5. What this means for enterprise buyers

The three questions from the February edition still narrow the field. The directory now answers them.

1. Where does your customer data live? If the warehouse is the system of record, start with the reverse ETL platforms and the packaged and suite platforms above 20 percent (Salesforce Data 360, Treasure AI, Redpoint, Simon Data). If the CDP will be the profile store, the warehouse share matters less than the inbound catalog.
2. Which direction do you need most? Match the In/Out profile. An ingestion-heavy catalog (Salesforce Data 360, Klaviyo, Braze, Treasure AI) is built to assemble the profile; an outbound-heavy one (Tealium, mParticle, Hightouch, Zeotap, Zeta) is built to activate it.
3. Which specific tools must connect? Use the directory's tool view: of 2,365 tools, 70 percent are documented by one platform. Shortlist by the tools in your stack, then compare the three or four platforms that document all of them.

Each figure in this report links to the matching directory filter, so the comparison can be rerun as catalogs change.

How we count

• Scope: 44 customer data and engagement platforms; every integration with a third-party tool that the platform documents on a public page. The platform's own SDKs, APIs, tags and features are not counted.
• Source: each listing links to the page it came from. Descriptions are CdpHub's, written from that page; a listing's direction, category and run type come from the page.
• Listings and tools: a platform with several documented integrations for one tool (audiences and conversions, source and destination) has several listings and one tool. Renamed products are grouped under the current name.
• Run types: client-side = a tag, pixel, SDK or plugin in the site or app carries the data; server-side = the platform's servers exchange data by API, webhook, stream or file; warehouse = the tool is a warehouse, lake, bucket or database, or the integration is reverse ETL, zero-copy or data sharing. Reverse ETL integrations are counted as warehouse only.
• Every page was read within its site's robots.txt rules. ActionIQ's catalog listings have no run type, and SAP CDP documents its integrations on a help portal those rules close to our reviewers; both are marked in the table.
• Counts are as of October 2026 and change as catalogs change; the live directory is the reference.