The shopper data platform (CDP) industry is entering a brand new phase as standalone CDPs give solution to broader enterprise marketing suites.
Latest platform enhancements from SAS, an analytics and AI software firm, reflect the growing integration of customer data management with enterprise marketing platforms.
CDPs serve because the central data engine for commerce and CRM, cleansing and stitching together data from website activity, mobile apps, emails, and point-of-sale systems — including anonymous browsing behavior, online orders, and offline store visits — to create a unified, 360-degree profile of every customer.
Demand for unified customer data continues to grow, but organizations increasingly expect those functions to be built into broader cloud ecosystems and native martech suites.
The Gartner 2025 Magic Quadrant for Customer Data Platforms predicts that by 2028, the info management market will converge right into a single data ecosystem enabled by data fabric and GenAI.
Lisa Loftis, principal management consultant of customer intelligence at SAS, sees the corporate’s embedded, composable CDP inside Customer Intelligence 360 as an approach with long-term endurance. As a substitute of traditional CDPs that require organizations to repeat customer data into one system, SAS lets marketers activate data directly from their cloud data stores.
“At the very least for those CDP vendors whose eyes have all the time been on the final word prize — native architecture from the bottom up across all customer engagement capabilities, it provides true composability in each data and features and real-time decisioning that is really real-time,” she told the E-Commerce Times.
She sees growing agreement amongst marketing technology analysts that CDPs will grow to be standard components of enormous enterprise marketing solution suites, reminiscent of customer engagement platforms or real-time interaction engines.
Standalone CDPs Are Giving Approach to Enterprise Suites
Loftis believes standalone CDPs are disappearing as composable architectures grow to be the popular approach, with CDP functionality becoming an embedded feature inside broader enterprise suites. She expects independent vendors to face increasing pressure unless they’re acquired or broaden their capabilities.
“As more firms buy into the cloud data warehouses and place more emphasis on comprehensive capabilities around journey orchestration, AI-driven insights, and real-time decisioning, CDPs with limited capabilities will proceed to lose their luster,” she said.
Loftis noted that Gartner’s latest CDP evaluation indicates the industry is moving toward two broad approaches: enterprise customer engagement platforms and AI-driven capabilities layered on top of cloud data warehouses.
She added that Gartner expects customer data decisions to increasingly involve marketing, sales, finance, supply chain, and customer support, reflecting the expanding enterprise role of customer data.
Composable Data Comes With Hidden Pain Points
Loftis said composability is usually about each data strategy and capability modularity, which might introduce significant issues on each side. One issue is that CDP activation is simply as effective as the standard of the underlying cloud data warehouse. Organizations still must make sure the underlying data is clean.
“This isn’t a given, and it isn’t something typically addressed by vendors selling zero-copy CDP tools, thus becoming a problem discovered after the CDP is purchased. And addressing these issues comes at a value that can be not factored into the worth of the CDP itself,” she explained.
Compute costs could be one other significant factor. Cloud data warehouse vendors charge based on compute usage. When typical CDP activities — identity resolution, audience constructing, analytics — occur directly in a cloud data warehouse, the organization incurs additional compute costs.
“CDP-related workloads can significantly increase cloud data warehouse usage due to the frequency and breadth of the queries required to keep up quality,” Loftis added.
Loftis also cited personally identifiable information (PII) as a priority tied more to the modular architecture of composable CDPs than to data strategy itself. While most customer data stays within the cloud data warehouse, PII still must move to the tools and channels that handle journey orchestration, decisioning, and activation. In a more robust CDP with those capabilities inbuilt natively, that becomes less of a problem.
“Nonetheless, if each capability is carried out in a unique toolset or application, the PII could be duplicated widely,” she warned.
Automated Decisioning Needs Higher Guardrails
Loftis agreed that the growing use of automated decision-making engines that act on real-time customer data requires brands to determine recent guardrails. These should include explainability and transparency so marketers can understand decision logic, AI model behavior, and outcomes.
GenAI should provide this information in clear, concise language that marketers can understand while not having a knowledge science background. It also needs to analyze the choices, make recommendations for changes based on decision performance and model health, and account for considerations reminiscent of contact policies and arbitration between competing offers, she suggested.
“Models in the choice intelligence arena also needs to include automatic bias mitigation and be retrained as needed, ensuring trustworthy and responsible AI-powered decisions,” she said.
Loftis added that her recommendations are all a part of human-in-the-loop initiatives. If agents are used to make decisions, a human must have ultimate responsibility for approving the ultimate decision before it’s deployed.
“There also needs to be a two-way data flow from the choice execution back to the choice intelligence environment for decision results in order that performance could be analyzed and decisions and models refined repeatedly,” she added.
Has the CDP Outgrown Its Category?
The emphasis isn’t any longer on unifying customer data but on activating it natively inside broader enterprise platforms. Loftis noted that CDPs were originally designed to assist marketers unify, segment, and activate customer data. The goal was to resolve the challenge of integrating customer data from disconnected systems.
“The category has evolved to focus as much on what marketers do with the info as on getting it integrated and unified,” she reasoned.
She pointed to comments by David Raab, founder and CEO of the Customer Data Platform Institute, who argued last 12 months that the CDP category is evolving reasonably than disappearing.
“There’s nothing radical a few CDP being embedded in a customer-facing system. In truth, statistics in our Industry Update report have long shown that what we call ‘campaign’ and ‘delivery’ CDPs comprise greater than two-thirds of the industry. The reality is, the market way back decided it preferred a CDP that was part of a bigger product. So the most recent round of acquisitions reflects a continuation of that situation, not a radical departure,” she said.