How to calculate the value of your company's data
Most senior leaders know their data is worth something. Very few can say what. That gap - between intuition and number - is where bad decisions live.
The mistake organisations at this scale make most often is treating data as an IT asset rather than a commercial one. It sits on an infrastructure budget, governed by a data team, reviewed in a technology steering committee. No one is asking what it would cost to not have it. No one is modelling what a competitor could do if they had it instead. No one is putting a figure next to it on the balance sheet.
This article will not tell you that your data is your most valuable asset. You have heard that. It will show you how to actually calculate what it is worth - using methods that hold up in a boardroom, a due diligence process or a strategic planning cycle - and it will tell you what to do when the number surprises you.
Why data valuation is not an academic exercise
If you are a CFO at a firm turning over £800m, you are already making decisions that implicitly value your data. You are deciding whether to invest in a data platform. You are deciding whether to hire a CDO. You are deciding whether to migrate to a new CRM or ERP. Every one of those decisions carries an assumption about what the underlying data is worth - you are just not making that assumption explicit.
That matters for three reasons.
First, without a valuation you cannot prioritise. You will spread investment across every data initiative that generates internal enthusiasm rather than concentrating it on the domains where data creates the most commercial value.
Second, private equity and acquirers are increasingly pricing data capability into transactions. If you cannot articulate the value of your data in a structured way, you will leave money on the table in any M&A context - whether you are the target or conducting due diligence on a bolt-on.
Third, it changes the conversation at board level. A CFO who can say "our customer transaction dataset has an estimated replacement cost of £12m and generates an estimated £4m per year in margin through personalisation" is having a different conversation to one who says "data is strategically important."
The three methods that actually work
There is no single universally accepted method for data valuation - which is both the problem and the opportunity. The right approach depends on why you are valuing the data. Use the wrong lens and you will either undervalue or overstate, both of which carry risk.
Cost-based valuation asks: what would it cost to recreate this dataset from scratch? This is the most defensible method in a due diligence context. Include data collection costs, cleaning and normalisation, storage, licensing (if third-party data is involved) and the time cost of the internal teams who built and maintain it. A mid-size retailer might find that its twelve years of transaction-level customer data, properly costed, represents £6m to £10m in replacement value once you account for the irreplaceable time dimension - you simply cannot buy twelve years of purchase history.
Income-based valuation asks: what economic benefit does this data generate, or what would it generate in a more capable state? This is the most useful method for internal prioritisation. Take a specific use case - say, a logistics business that uses route and demand data to reduce empty miles - and model the margin impact. If the data enables decisions that save £3m per year in operating cost, discount that over a useful life of five years at an appropriate rate, and you have a defensible income-based value.
Market-based valuation asks: what would a third party pay for this data? This is relevant when you are considering data licensing, a data joint venture or a sale. It is the hardest to calculate directly but the most commercially motivating. A financial services firm with proprietary transaction data across a specific sector may find that a data broker or research firm would pay materially for access - which means the firm is effectively sitting on a revenue line it has not opened.
The most rigorous approach uses all three and triangulates. Where they converge, you have a credible figure. Where they diverge, the gap itself is informative.
A practical framework for getting started
You do not need to value every dataset. You need to value the ones that drive commercial decisions.
Start by mapping your data domains against four criteria:
- Exclusivity - can a competitor buy this data elsewhere, or is it proprietary to your business?
- Depth - how much history does it cover, and at what granularity?
- Linkage - can it be combined with other datasets to generate insight that neither could produce alone?
- Activation - is it currently being used in decisions, and at what frequency?
Score each domain against these four criteria. The datasets that score highest on exclusivity and linkage are almost always the ones most undervalued on the balance sheet and most relevant to competitive advantage.
A B2B professional services firm running this exercise might discover that its proprietary benchmarking data - collected across ten years of client engagements - scores highly on all four dimensions but is being used only for internal reporting. That is a valuation gap. It is also a product opportunity.
Once you have identified your priority domains, apply the three methods above to each one. The output is not a precise number - it is a range, a set of assumptions and a view on which assumptions matter most. That is enough to make a decision.
What to do when the number is larger than you expected
Most organisations that run this exercise for the first time arrive at a number that is uncomfortably large relative to the investment they have made in data infrastructure, governance and capability.
That discomfort is useful. It means you have been underinvesting relative to the value at stake.
The practical response is to build a data value realisation plan. This is not a data strategy - those tend to be long on vision and short on commercial specificity. A value realisation plan starts from the valuation outputs, identifies the specific decisions or products that each dataset enables, and assigns ownership and investment accordingly.
For a £1bn ecommerce business, this might mean recognising that its behavioural clickstream data - currently used only for retargeting - has significant income-based value in demand forecasting, supplier negotiation and dynamic pricing. Each of those represents a separate initiative with a modellable return. Tools like Quantsole can accelerate this by making that data queryable in natural language across commercial teams who would not otherwise access it - reducing the time between insight and decision.
The other situation worth naming is M&A. If you are a PE-backed business preparing for exit, or a firm conducting tech due diligence on an acquisition target, a data valuation exercise done now is worth considerably more than one done under time pressure during a process. Buyers who understand data are pricing it. Sellers who do not will be priced against.
Start with one domain, not the whole estate
The instinct at this scale is to want a complete picture before committing to anything. That instinct will stall you.
Pick one dataset that you already suspect is undervalued. Apply the three valuation methods. Run the four-criteria scoring. Produce a number and a set of assumptions. Present it to the CFO or the board alongside the investment required to fully activate it.
That one exercise will do more to shift the conversation around data than any strategy document. It puts a commercial figure next to an asset that has been treated as a cost, and it forces a decision: invest to realise the value, or accept that you are leaving it on the table.
Organisations that do not do this are not standing still. Their competitors - including PE-backed challengers with mandated data transformation programmes - are building the capability to extract value from data that your business has not even identified yet. The cost of delay is not hypothetical. It is measurable. And it compounds.
If you want to run a structured data valuation exercise for your business, Rodan offers a diagnostic engagement that produces a prioritised domain map, valuation estimates and a clear view of where the commercial opportunity is largest. It takes two to four weeks and is designed to give you a decision, not a report.
[Book a diagnostic with Rodan at rodan.io]
