Somewhere inside a large company, Don Bulmer exists six times. Salesforce knows Don the prospect. The billing system knows Donald the account contact. A webinar platform remembers a work email he stopped using. Another database has the right person and the wrong company. This is not a hypothetical nuisance. Let those records drift and a salesperson calls twice, a privacy request misses half the profile, a rebate goes to the wrong account, or an AI agent makes a confident decision about a customer who does not quite exist.
Reltio is the software referee in that argument. It pulls records from operational systems, warehouses, documents and third-party feeds; determines which ones describe the same real-world entity; applies rules about which values survive; preserves where the data came from; and serves the resulting profile back to applications in real time. The industry calls the result a golden record. Reltio now prefers the broader phrase system of context, because a name and address are less useful than the relationships around them: this person belongs to that household, buys this product, works for that company and has granted these permissions.
The 10-second product tour
The first thing to break was the old stack
Founder Manish Sood spotted the opening in 2011. Big-data infrastructure was becoming practical, enterprises were warming to software as a service, and business data was splintering across an accelerating number of applications. Sood had worked at Informatica and Siperian, so this was not a tourist's observation. His bet was architectural: master data management should be multitenant, API-first and born in the cloud, not an on-premise product wearing a hosting costume.
Life sciences supplied the first hard test. Drug companies manage dense webs of physicians, hospitals, products, locations, consent, clinical trials and regulation. The data changes constantly, and the cost of a false match is higher than sending someone the same coupon twice. Reltio proved its multidomain model there, then carried the pattern into healthcare, financial services, insurance, retail, manufacturing, technology, travel and hospitality.
The clearest proof is AstraZeneca. Its estate had swollen to 67 master-data deployments across more than 140 countries. Upgrades were costly, sales searches were slow, and patient consent was difficult to keep synchronized. The company replaced those deployments with three interlinked regional Reltio hubs. AstraZeneca says the move saved £3.6 million a year in infrastructure and management costs, lifted sales-rep productivity tenfold and put a patient master live in 14 weeks.
“We couldn't deliver [a 360-degree customer view] without having the data, systems, and technology together in a connected format.”Shamim Mohammad, CarMax
What customers are actually buying
The core product is Reltio Data Cloud, now presented as the Reltio Context Intelligence Platform. Multidomain MDM handles the old but essential work: identity resolution, match and merge, hierarchies, reference data, survivorship, stewardship, quality and lineage. Intelligent 360 turns mastered entities into reusable customer, product or supplier views. An integration layer moves them to operational systems, analytics platforms and models.
The differentiation is speed plus shape. Traditional MDM often centered on batch jobs and rigid relational models. Reltio models entities and their relationships as a graph, exposes them through real-time APIs, and runs as a multitenant SaaS service across major clouds. Industry “velocity packs” bundle data models, match rules, interfaces and integrations learned from earlier projects. That packaging matters. Most expensive enterprise software does not fail because a button is missing; it fails while a committee debates what “customer” means.
The newer products aim at AI workloads. FERN uses pretrained models to resolve identities without requiring teams to hand-author every matching rule. AgentFlow automates repetitive stewardship jobs such as resolving possible duplicates, profiling incoming data, spotting anomalies and assigning reviews. Lightspeed is a data-delivery network that Reltio says can respond in under 50 milliseconds. Unstructured Data Studio extracts people, organizations and products from PDFs, transcripts and web content, then connects them to governed records.
What does it cost? Reltio does not publish enterprise list prices. A buyer's bill depends on profile volume, domains, API usage, environments, integrations, resilience and support, then adds implementation and the human cost of repairing source data. That makes the honest comparison total cost over several years, including the legacy systems and manual work a company can retire. AstraZeneca's savings are persuasive because they count that whole estate, not a cheap-looking license in isolation.
Snowflake and Databricks integrations reveal the larger strategy. Reltio does not need to become the warehouse. Zero-copy and data-sharing features let teams combine mastered profiles with current behavioral data where it already lives. Microsoft Fabric, Veeva Vault CRM, Salesforce, Kafka and more than 1,000 prebuilt connectors extend the same idea: become the trusted middle, not the only database in the building.
| Choice | Best when | The catch |
|---|---|---|
| Reltio | Complex, real-time, multidomain data across mixed cloud and application estates | Enterprise scope, private pricing and serious governance work |
| Informatica, Semarchy, Profisee, Stibo | A different ecosystem, deployment model or domain specialty fits better | Every vendor makes different tradeoffs in flexibility, operations and speed |
| Build around a warehouse | A skilled team has narrow use cases and mostly analytical needs | Operational write-back, identity rules, lineage and stewardship become your product |
The $237 million runway ends at SAP
Reltio financed the long climb with five main rounds: $10 million in 2015, $22 million in 2016, $40 million in 2017, $45 million in 2018 and $120 million in 2021. The last round, led by Brighton Park Capital, brought total funding to $237 million and valued the company at about $1.7 billion. It crossed $100 million in annual recurring revenue in 2022.
Reported annual recurring revenue
By the fiscal year ended January 31, 2026, Reltio reported $185 million in ARR, 40 percent year-over-year bookings growth, positive cash flow and more than 200 large-enterprise customers. Forty-nine customers were spending more than $1 million annually; 43 were in the Fortune 500. Those are company-reported figures, but they explain why an established software buyer paid attention.
SAP announced the acquisition in March and closed it on May 7, 2026. The price was not disclosed. The logic was. SAP wants its Business Data Cloud and Joule agents to work with both SAP and non-SAP information. Reltio can identify and govern the customers, products, suppliers and employees those agents reason about. SAP said Reltio would become a core capability in the Business Data Cloud while remaining available as a standalone product for the foreseeable future.
This is where the story acquires useful tension. Reltio has sold openness across heterogeneous technology stacks. SAP owns one of the largest application estates on earth. The acquisition creates distribution and deeper integration, but buyers will watch whether non-SAP connectors, cloud choice, partner economics and standalone pricing remain first-class. “Open” is easy to print on launch day. It is tested in roadmaps.
The playbook worth stealing
First, start where the pain is severe enough to pay for rigor. Life sciences forced Reltio to handle regulation, identity and many data domains early. Second, turn repeated services work into product defaults. Velocity packs compress years of implementation decisions into a starting schema and workflow. Third, sell the before-and-after architecture. “Context intelligence” is abstract; 67 systems becoming three is legible in a budget meeting.
Fourth, fit into the stack customers already have. Reltio wins when it makes Snowflake, Databricks, Veeva, Salesforce, SAP and operational apps more reliable, not when it demands that all of them disappear. Finally, connect AI to a dull, expensive workflow. A match-resolution agent is less theatrical than a chatbot. It may also save a data steward hours of repetitive review.
When this does not work
- A small business has only a few clean source systems and no multidomain identity problem.
- No executive owns the definitions, merge policy or remediation work.
- The organization wants a one-time cleanup instead of continuous governance.
- Source systems keep producing bad records and nobody is authorized to fix them.
- A neutral independent vendor is a hard requirement after the SAP acquisition.
Master data management is not a magic eraser. Software can suggest that two records match; someone still needs to decide what counts as evidence, who may override the result and where corrections flow. Without those decisions, a new platform merely creates a better-lit room for the same argument.
Reltio's achievement is making that room operational at enterprise scale. The company did not invent duplicate customers or corporate politics. It built a cloud service that can reconcile the first while exposing the second. In the age of AI agents, that formerly boring job has become a prerequisite. SAP did not disclose what it paid. The strategic price of skipping the data layer is easier to see.