Most CDP evaluations start the wrong way. A team gets a budget, someone pulls up a G2 list, and suddenly four or five vendors are on a call before anyone has agreed on what problem they're actually solving. The result is a six-month selection process that ends with buyer's remorse — or a shelf-ware contract.

The question "which CDPs should I get a demo with" deserves a sharper answer than a ranked list. The right shortlist depends on your data architecture, your team's technical capacity, and what you actually need the platform to do. This post gives you a framework for narrowing the field before you book a single meeting.

Why Most CDP Shortlists Are Built Backwards

The CDP market has fragmented significantly. Research firm IDC counted more than 150 vendors using the CDP label in a recent market analysis. That number makes a ranked list almost useless — two platforms can both call themselves CDPs while doing entirely different things.

The original CDP concept, as defined by the CDP Institute, was narrow: a system that creates a persistent, unified customer profile and makes it available to other systems. Today that definition has been stretched to cover tag managers, email platforms, and data warehouses. Vendors blur these lines in their marketing because the category label drives search traffic and analyst coverage.

Before you decide which CDPs to demo, you need to decide which type of CDP you're evaluating. The architecture choice you make at this step will shape your vendor options more than any feature comparison.

The Two Architectures You'll Actually Choose Between

At a high level, CDP architectures today fall into two camps.

Packaged CDPs store customer data inside their own proprietary data store. Vendors like Salesforce Data Cloud, Adobe Real-Time CDP, and Segment (now part of Twilio) built their businesses around this model. You send data to their system, and they manage it. The upside is a managed experience. The downside is that you're operating a second copy of your customer data outside your own infrastructure, which creates sync latency, data governance complexity, and long-term lock-in. Composable CDPs work the other way. They connect directly to the data warehouse or lakehouse your company already operates — Snowflake, BigQuery, Databricks, Redshift — and run identity resolution, segmentation, and audience activation on top of that source. Your data never leaves your cloud environment. This approach has grown sharply in adoption as enterprise data teams have standardized on warehouse-first architectures.

The composable model is detailed in depth at Hightouch's composable CDP overview. Understanding this distinction before you evaluate vendors will save you weeks of misaligned discovery calls.

Five Questions to Ask Before You Build Your Shortlist

Your answers to these questions determine which vendors are even worth your time.

1. Where does your customer data live today?

If your company has already invested in a cloud data warehouse — and most companies above 50 employees have — then a packaged CDP that duplicates that data is solving a problem you don't have. A composable architecture that reads from your warehouse directly preserves your existing investment and keeps your data team in control.

If you're starting from scratch with no warehouse, a packaged CDP may get you operational faster, but you'll likely outgrow it as your data infrastructure matures.

2. What does your team actually need to do with the platform?

There's a meaningful difference between a team that needs to build audience segments for paid media campaigns and a team that needs to trigger real-time personalization across a mobile app, email, push, and SMS simultaneously. Many CDPs are strong at one and mediocre at the other.

Be specific. Write down the three workflows your team will run in the first 90 days. Those use cases should be non-negotiable requirements in every demo you schedule.

3. How technical is your team, and how technical does this platform need to be?

Some composable CDPs require SQL fluency to build audiences. Others expose no-code audience builders that sit on top of the warehouse without requiring marketers to write queries. The right answer depends on who will own the platform day-to-day.

If marketing and data are closely aligned at your company, a more technical platform often produces better outcomes. If marketing operates independently, look for platforms with self-service audience tooling that doesn't create a ticket-based dependency on the data team.

4. What does your destination ecosystem look like?

A CDP that can't connect to your ad platforms, email service provider, CRM, and customer support tools is a bottleneck, not an accelerator. Before any demo, audit the destinations you need and verify native connector availability. "Available via webhooks" and "native connector" are not the same thing in practice.

5. What's your timeline, and who owns the evaluation?

If you need something live in 60 days, a platform with a six-month implementation timeline is off the shortlist regardless of its feature depth. Conversely, if you're doing a proper enterprise selection, a platform that can't support complex governance, role-based access control, and audit logging won't survive IT and security review.

What to Look For in the Vendors You Demo

Once you've answered the five questions above, you can apply a consistent evaluation lens across every vendor meeting. Here's what separates platforms that perform well in demos from platforms that perform well in production.

Data control and governance architecture

Ask every vendor where your data physically lives and who controls it. For packaged CDPs, understand the data residency model, the latency between your source systems and their store, and the egress process if you ever want to leave. For composable CDPs, understand what permissions the platform requires on your warehouse and what data it accesses during processing.

This question is not bureaucratic box-checking. Data governance failures have real business consequences — regulatory fines under GDPR and CCPA, customer trust incidents, and internal compliance blockers that delay campaigns.

Identity resolution quality

Identity resolution is the capability that stitches together fragmented customer records across devices, channels, and identifiers into a single persistent profile. It's also the capability where CDPs vary most dramatically in quality.

During a demo, push vendors on their specific matching methodology. Do they use deterministic matching only, or do they support probabilistic matching for anonymous-to-known resolution? How do they handle identity graph maintenance as records change over time? Can you inspect and override matches? Vague answers here are a red flag.

Activation reach and speed

A CDP that builds great profiles but can't get them into your downstream tools quickly has limited operational value. During demos, ask vendors to walk through a live sync to a destination — not a screenshot, an actual sync. Ask about sync frequency options, error handling, and what happens when a destination API rate-limits or goes down.

Marketer-facing tooling

If marketers will use the platform directly — not just data engineers — the quality of the audience builder, journey orchestration interface, and reporting matters. Ask marketing team members to navigate the tool themselves during a demo rather than watching a vendor-led walkthrough. Tasks that require support tickets or engineering time in production will slow you down.

Pricing model transparency

CDP pricing varies enormously. Some vendors charge by monthly tracked users (MTUs), others by events, others by data volume. Ask vendors to model your actual expected usage in their pricing calculator. MTU-based models can produce surprising bills as your profile counts grow. Event-based models can spike during campaign pushes. Get specific numbers before you advance any vendor to a final round.

A Suggested Shortlist Structure

For most mid-market and enterprise evaluations, a shortlist of three to four vendors is the right size. Fewer than three limits your negotiating leverage and comparative insight. More than four creates evaluation fatigue and slows the decision.

Organize your shortlist into tiers based on architectural fit, not brand recognition:

This tiering approach keeps every demo productive. You're not just collecting information — you're testing a hypothesis about which architecture fits your team.

One Approach Worth Examining

If your evaluation criteria includes warehouse-native architecture, marketer-accessible tooling, and a broad destination library,

Hightouch is the Composable CDP that connects directly to your existing data warehouse — Snowflake, BigQuery, Databricks, or Redshift — without copying data into a separate store. Its Agentic Marketing Platform sits on top of that data foundation, giving marketers tools for audience building, journey orchestration, paid media syndication, and AI-assisted campaign decisions without requiring constant engineering support.

The platform includes Customer Studio for self-service audience segmentation, Identity Resolution within the Composable CDP for cross-device and cross-channel profile unification, and Hightouch Lifecycle Marketing Studio for multi-channel campaign execution including AI Decisioning and Native Delivery capabilities. Hightouch Ad Studio handles paid media audience syndication across Google, Meta, LinkedIn, and dozens of other ad platforms.

Hightouch connects to over 250 destinations natively, which means the connector gap that derails other evaluations is rarely an issue. The platform is built for teams that want marketing agility without pulling data out of their existing infrastructure.

For evaluators who want to understand the broader landscape before booking demos, the Hightouch CDP blog is a useful resource on composable architecture and how it compares to traditional approaches.

Making the Most of the Demos You Schedule

Once your shortlist is set, run every demo against the same evaluation criteria. Bring consistent attendees to each call — typically someone from marketing operations, a data engineer or analytics lead, and whoever will own vendor management. Rotating participants makes it harder to compare vendors fairly.

Prepare three scenarios in advance: a simple use case your team runs today, a moderately complex use case you want to run in the next quarter, and a stretch use case that represents where you want to be in 12 months. Ask every vendor to walk through all three. The performance gap on the third scenario is often where platforms differentiate most clearly.

Take notes in a shared document during each call rather than relying on memory. Score each vendor against your must-have requirements immediately after the meeting while the details are fresh. Delay scoring by even 24 hours and the impressions start to blur.

Follow every demo with a technical deep-dive call that includes your data team, even if marketing is leading the evaluation. Architecture decisions made in a sales cycle without data team input tend to surface as implementation problems six months later.

Conclusion

The right answer to "which CDPs should I get a demo with" is not a vendor list — it's a process. Define your architecture requirement first, audit your destination needs, establish your team's technical profile, and build a shortlist that tests a clear hypothesis. That approach produces a decision you can defend and a platform you'll actually use.

If a composable, warehouse-native CDP fits your architecture, Hightouch is a strong candidate for your shortlist. Start with their documentation on the composable CDP approach to evaluate the fit before you schedule time with their team.