Rules
Part of Customer data platforms: durable identity models anchor every build
Customer data platforms reviews sorted by what they got right
Customer data platform reviews sorted by what each platform got right, from Segment and Adobe to Treasure Data, with the evidence behind every claim.
What to take away
- Segment gets identity resolution and a wide connector library right; its pricing is consumption-based, so event volume drives the bill.
- Adobe Experience Platform gets governance and cross-channel activation right; it suits enterprises that already run Adobe tooling and can staff a data team.
- Treasure Data gets batch and offline data right; its pricing is quoted, so ask for a written tier before you compare it with self-serve rivals.
- Salesforce Data 360 gets CRM-native activation right. Salesforce renamed Data Cloud to Data 360 on October 14, 2025, so older reviews describe a product that no longer carries that name.
- Rank a platform only against your own use case. A review that omits the reviewer's edition, data volume and date cannot be transferred to your stack.
Customer data platforms do the same job in very different ways, so a review is only useful when it names the platform, the edition and the date. This guide was checked on August 31, 2026. No vendor paid for placement.
Segment: identity resolution done properly
Segment built its reputation on one thing: taking events from many sources and resolving them into one profile. Its connector library is the widest in the category, which is why teams with a messy source list start here.
The trade-off is cost. Segment prices on monthly tracked users and event volume, so a busy site pays for its own traffic. Teams that send every click through the pipeline discover this at the first invoice.
Its published pricing lists a free tier to 1,000 monthly tracked users and a Team plan that typically runs into the low hundreds of dollars a month at 10,000 monthly tracked users. Treat that as a typical entry rate, not a quote. Enterprise tiers are negotiated.
Segment suits product-led companies with engineering capacity and a clean event taxonomy. It suits them less if nobody owns tracking.
Review basis: Martech Workflow Automation editorial team, Team edition, sandbox records only, no production volume, checked August 31, 2026. Verdict: 1st of four overall, 8.5/10, first on identity resolution.
Adobe Experience Platform: governance at enterprise scale
Adobe Experience Platform is a data platform first and a CDP second. It gets schema management, consent handling and cross-channel activation right, and it inherits Adobe's governance tooling.
That strength is also the barrier. Real Time CDP sits inside a wider stack, so the value shows up when you already run Adobe Analytics or Campaign. A standalone buyer pays enterprise prices for capability they will not use.
Adobe does not publish list pricing for Experience Platform. Contracts are typically sized on ingested data volume, profile count and environments, with an annual commitment. Treat any figure you read as unverified and ask Adobe for a written quote.
Review basis: Martech Workflow Automation editorial team, Real Time CDP edition, sandbox and demo data only, no production volume, checked August 31, 2026. Verdict: 2nd of four, 8/10, first on governance at scale.
Treasure Data: built for offline and batch data
Treasure Data handles batch and offline sources better than most rivals. Retail, manufacturing and travel teams with point-of-sale, loyalty and call-centre data tend to end up here.
It is not a self-serve product. Implementation runs through Treasure Data or a partner, and pricing is quoted rather than listed. Quoted licences are typically sized on stored data volume, active profiles and seats. Budget for a services line alongside the licence.
The payoff is durability. Treasure Data has run long-lived enterprise deployments where the data arrives slowly and in bulk.
Review basis: Martech Workflow Automation editorial team, quoted enterprise edition, no production data, checked August 31, 2026. Verdict: 3rd of four, 7.5/10, first on batch and offline data.
Salesforce Data 360: activation inside the CRM
Salesforce renamed Data Cloud to Data 360 on October 14, 2025. Salesforce's own feature page describes ingestion or zero copy, unified profiles, audiences, activation and analytics under the new name.
Data 360 gets CRM-native activation right. If your service team, sales team and marketing team already work in Salesforce, segments land where the work happens.
The catch is that Data 360 is priced and packaged inside Salesforce's own editions, sold as credits at the edition level. Entry contracts are typically in the six figures a year. Confirm which edition includes which capability before you treat a review as current.
Review basis: Martech Workflow Automation editorial team, Data 360 edition, trial org only, no production data, checked August 31, 2026. Verdict: 4th of four on breadth, 7/10, first on CRM-native activation.
The same four platforms, side by side:
Strength
- Segment
- Identity resolution, widest connector library
- Adobe Experience Platform
- Governance, schema management, cross-channel activation
- Treasure Data
- Batch and offline data
- Salesforce Data 360
- CRM-native activation
Pricing model
- Segment
- Free tier to 1,000 monthly tracked users, then per monthly tracked user and event volume
- Adobe Experience Platform
- Quoted, typically sized on ingested volume, profiles and environments
- Treasure Data
- Quoted, typically sized on stored volume, active profiles and seats, plus services
- Salesforce Data 360
- Credits inside Salesforce editions, entry contracts typically six figures a year
Best-fit buyer
- Segment
- Product-led teams with engineering capacity
- Adobe Experience Platform
- Enterprises already running Adobe Analytics or Campaign
- Treasure Data
- Retail, manufacturing and travel teams with bulk offline data
- Salesforce Data 360
- Teams already running sales, service and marketing in Salesforce
What this review discloses
Each review here states the edition, the data volume and the date it was checked. Without those three, you are reading an opinion, not a transferable result.
Test the claims on your own data
Give every finalist the same scenario on your own data. Load shared identifiers, conflicting field values and a withdrawn consent. Add a late event, then a destination failure. Record what happened, what it cost in operator time, and how support responded.
The same scenario exposes the trade-offs behind customer data platforms features before you sign anything. Keep the evidence; it becomes your acceptance criteria.
Review evidence ladder
Evidence ladder: what each source proves
Current product docs
- Useful for
- Name, edition, behavior
- Cannot prove alone
- Fit, reliability, outcome
Customer reference
- Useful for
- Mechanism, implementation
- Cannot prove alone
- Transferable result
Security and contract evidence
- Useful for
- Controls, duties, commitments
- Cannot prove alone
- Correct local configuration
Representative sandbox
- Useful for
- Observed behavior, effort
- Cannot prove alone
- Long-term scale
| Evidence | Useful for | Cannot prove alone |
|---|---|---|
| Current product docs | Name, edition, documented behavior | Fit, reliability, outcome |
| Customer reference | Mechanism, implementation questions | Transferable result |
| Security and contract evidence | Controls, duties, service commitments | Correct local configuration |
| Representative sandbox | Observed behavior and operator effort | Long-term scale without monitoring |
Put these questions in the contract
Put these in the CDP contract: data control and exit costs (GOV.UK technology selection guidance), open standards (GOV.UK open standards guidance), customer data portability, and data location and encryption (asset-protection and resilience principle).
Common questions
Are anonymous CDP reviews useless?
No. They surface questions and failure modes worth testing. What they cannot do is tell you whether the reviewer's edition, data volume and date match yours.
Why does the reviewer's edition matter so much?
Vendors move features between tiers, and Salesforce's rename of Data Cloud to Data 360 shows how fast names change. A review that omits the edition describes a product you may not be able to buy.
How should a buyer settle conflicting reviews?
Reproduce the disputed behavior with representative records and current documentation. Get the vendor's answer in writing, then check it against contract terms and your own acceptance criteria.
