Data assets are increasingly recognised as among the most valuable intangible assets in modern businesses. Under IFRS 3, databases and data compilations qualify as technology-related intangible assets when they meet the separability or contractual-legal recognition criteria. In practice, most proprietary databases are separable — they could be sold, licensed, or otherwise transferred independently of the business.
Yet data asset valuation remains one of the least standardised areas in intangible asset practice. Unlike customer relationships (where MPEEM is standard) or trade names (where RFR dominates), data assets do not have a single preferred method. The right approach depends on how the data creates value for the business.
Data assets are fundamentally different from other intangible assets because they are non-rivalrous (use by one party does not diminish them), they can be replicated at near-zero marginal cost, and their value often increases with scale. These characteristics mean that standard valuation frameworks must be adapted rather than applied mechanically.
Before selecting a valuation method, it is essential to understand the economic characteristics that distinguish data from other intangibles:
| Characteristic | Data Assets | Traditional Intangibles |
|---|---|---|
| Rivalrousness | Non-rivalrous — can be used simultaneously by multiple parties | Generally rivalrous |
| Marginal cost of use | Near zero | Varies |
| Network effects | Value often increases with volume | Less common |
| Perishability | Some data depreciates rapidly (market data); some is permanent (geological data) | Predictable depreciation patterns |
| Replicability | Can be copied perfectly | Cannot be perfectly replicated |
| Legal protection | Database rights, trade secrets, contractual restrictions | Patents, copyrights, trademarks |
| Combinability | Combining datasets can create value greater than the sum of parts | Limited combinability |
The cost approach values a data asset at the cost to recreate or replace it. This is the most commonly used method for data assets in purchase price allocations, particularly when the data's income contribution is difficult to isolate.
The cost of gathering equivalent data — surveys, sensors, web scraping, API purchases, manual data entry, or partnerships.
ETL (Extract, Transform, Load) processes, deduplication, normalisation, validation, and quality assurance.
Schema design, taxonomies, metadata creation, feature engineering, and integration with other datasets.
The economic value lost during the time it would take to rebuild the dataset — particularly significant for time-series data that cannot be retrospectively reconstructed.
The cost approach systematically undervalues data assets whose value derives from their use rather than their creation. A customer behavioural dataset that took £500K to compile might generate £5M in annual revenue improvements through personalisation algorithms. The cost approach captures the £500K; the income approach captures the £5M effect.
The income approach values a data asset based on the earnings it generates. This requires isolating the data's contribution to revenue — which is often the most challenging aspect of data asset valuation.
| Method | Application | When to Use |
|---|---|---|
| Relief from Royalty | Data asset could be licensed | When comparable data licensing deals exist |
| With-and-Without | Data improves business performance | When the data effect on revenue or cost is measurable |
| MPEEM | Data is the primary earnings driver | Rare — most data assets are contributory, not primary |
| Incremental cash flow | Data creates identifiable incremental revenue | When A/B tests or natural experiments can quantify the effect |
If the data asset is the type that is commonly licensed (market data, consumer demographic data, industry benchmarks), the RFR method can be applied. Data licensing royalty rates vary widely:
| Data Type | Typical Licensing Rate | Basis |
|---|---|---|
| Financial market data | 5-15% of revenue | Time-critical, high-value |
| Consumer demographic data | 2-8% of revenue | Volume-driven, competitive market |
| Industry benchmarking data | 10-25% of subscription revenue | Unique, high switching costs |
| Geospatial data | 3-10% of revenue | Specificity and freshness |
| Healthcare / clinical data | 5-20% of revenue | Regulatory barriers, scarcity |
The market approach looks for comparable transactions involving similar data assets. While a secondary market for data assets is emerging, it remains thin and fragmented. Sources of comparable data include:
The market approach is most useful as a cross-check rather than a primary method, given the difficulty of finding truly comparable data transactions.
Data asset useful lives vary dramatically depending on the type of data:
| Data Type | Typical Useful Life | Rationale |
|---|---|---|
| Real-time market data | Hours to days | Value expires with time |
| Customer contact data | 1-3 years | Contacts become stale |
| Customer behavioural data | 3-7 years | Patterns evolve gradually |
| Proprietary research data | 5-15 years | Slow obsolescence |
| Historical transaction data | 10-20 years | Training data for models |
| Geological / scientific data | 20+ years or indefinite | Permanent information |
Data assets often have a "compounding" useful life — the data becomes more valuable over time as the dataset grows, rather than less valuable through obsolescence. This is the opposite of most intangible assets and may justify an increasing (rather than declining) cash flow pattern in the valuation model.
Data privacy and GDPR. Personal data subject to GDPR or similar regulations cannot be freely transferred or licensed, which limits the data's separability and may reduce its fair value. The valuation must consider the cost of maintaining compliance and the risk of regulatory action.
Data quality and completeness. A database is only as valuable as the quality of its contents. Missing fields, outdated records, and inconsistent formatting all reduce the replacement cost and the income-generating potential.
Embedded vs standalone value. Many data assets are valuable only within the context of the business's technology platform and algorithms. A proprietary dataset that powers a machine learning model may have limited standalone value but enormous embedded value. The valuation must address which perspective is appropriate.
Data asset valuation is an evolving field that requires adapting traditional intangible valuation frameworks to the unique economics of data. Choose the cost approach when the data's income contribution is hard to isolate; use the income approach when the data demonstrably drives revenue or cost savings; and cross-check with market transactions where available. The Opagio Valuator identifies data assets as part of its comprehensive intangible asset assessment. Identify your data assets.
Ivan Gowan is the Founder and CEO of Opagio. His career in fintech included building data-driven trading platforms at IG Group, where proprietary market data and customer analytics were core strategic assets. Meet the team.
Estimate the value of your intangible assets using industry-standard methods like Relief from Royalty, MPEEM, and With & Without.
A practical guide to the six core methods for valuing intangible assets: Relief-from-Royalty, MPEEM, With-and-Without, Cost, Income, and Market approaches. Includes when to use each method and worked examples.
Read more →
How to value favourable contracts and order backlog as intangible assets in purchase price allocations. Covers the income approach for both asset types, the distinction between contractual and off-market benefits, and coordination with customer relationship valuations.
Read more →
How IP valuations work for lending — four accepted methods, IVS compliance requirements, what banks actually need, common failures, and how to get lending-ready.
Read more →Get the latest insights on intangible asset growth and productivity delivered to your inbox.
Book a free consultation to see how the Opagio Growth Platform can help your business.