Data as a Service
Data as a service is a business model in which a provider sells data itself — collected, cleaned, kept up to date and delivered on demand through an API, a feed or a cloud data share — usually for a subscription or usage fee. The customer buys the finished data rather than software to gather and process it themselves.
- Also known as
- DaaS, Data-as-a-Service
- Used at
- financial market data, credit and identity data, company and ownership records, location and property data, and training and evaluation data for AI models
- Delivery
- APIs for record-by-record lookups, bulk files or feeds, and direct shares into the customer's cloud data warehouse
- Pricing
- annual subscriptions by dataset, users or use case, per-call fees for lookups, or a mix; the licence limits what the customer may do with the data
- Rules
- personal data brings in the GDPR (applicable since 25 May 2018) and its UK equivalent; the EU Data Act (Regulation (EU) 2023/2854), applicable since 12 September 2025, gives users rights of access to the data their products generate
- Deals
- LSEG agreed in 2019 to buy the market-data business Refinitiv in an all-share deal valuing it at $27 billion; in June 2025 Meta paid more than $14 billion for a 49% non-voting stake in the AI training-data supplier Scale AI
- Other sense
- the term is also used for cloud tools that store, manage and analyse a customer's own data
How data as a service works in practice
The product is rarely the raw data, which is often public or cheap, but the work done on it: gathering from many sources, matching records that describe the same entity, correcting errors, filling gaps and keeping everything current. A company-data provider, for example, pulls filings from dozens of registries, works out that three spellings refer to one company, links it to its parent and pushes changes to customers within hours. Customers pay because doing that themselves would cost more.
Revenue is usually recurring, and gross margins can be high once the data is built, since serving one more customer costs little. The business is defensible when the data is hard to replicate — proprietary sources, data contributed by customers, long history — and when customers wire it into their own systems and models, which raises switching costs.
The weaknesses are on the supply side. If the data comes from a few licensed sources, those sources can raise prices, cut access or start selling direct. If it is scraped, the right to use it may be contested. If it identifies people, its collection and resale need a lawful basis.
For a buyer, diligence starts with rights: for each source, the licence, its term and any change-of-control clause; for each customer contract, what the customer may do with the data. Then retention, since subscription data businesses are valued largely on how reliably customers renew.
Example
A provider of UK and EU company-ownership data sells to banks and fintechs for know-your-business checks. It has 420 customers, £18 million of annual recurring revenue, net revenue retention of 112% and a gross margin of 78%; 70% of revenue comes from annual subscriptions and 30% from per-lookup API fees. A buyer's diligence finds that 35% of records depend on one commercial registry-data licence, which ends in 20 months and can be terminated on a change of control. The buyer makes a new licence with that source a condition of signing, and lowers its offer from 9× to 8× revenue to reflect the higher fee the source will charge — £18 million less for one contract.
Common mistakes
- Valuing the dataset without reading the licences behind it, some of which end on a change of control. - Selling data without terms that limit the customer's resale or use for model training. - Treating scraped data as owned data. - Booking usage-based fees as recurring revenue without checking how steady they are. - Overlooking data-protection rules when records identify people, not only companies. - Pricing a data business like a data-software business; the margins and risks differ.