Data valuation is the process of determining the monetary value of a dataset or collection of data. It’s one of the most fundamental processes in data commerce, but also beyond: with the increasing importance of data in today's economy, it's becoming crucial for organizations to understand the value of their data for reasons not restricted to financial gain. Data valuation can help organizations make better decisions about how to handle their data, including how much to invest in data security, how to prioritize data-related projects, and how to monetize their data.
While there is no established method for calculating data's value, there are several proposed methods, including the market-based model, economic model, and dimensional model. By understanding the value of their data, organizations can cultivate a data-orientated mindset across all departments, which has benefits going beyond the purely analytical.
On that, let’s look into the many reasons that data valuation is important in across types of organization and team in Part 1: why is data valuation important?
Only when you know the true value of an asset can you decide how to handle it. The same applies to data, which we know is valuable: intangible assets are now responsible for 90% of all business value. More businesses are realising that there’s value in their data. The logical next step is to specify and quantify that value by asking: at current market rates, how much is my data monetarily worth? However, there is no established method to calculate data’s value so empirically. As a report by Harvard Data Science Review puts it, ‘we as a society, want a way to value data in concrete terms. We are not there yet.’ Data valuation is process in its infancy - we know that it’s important, but we’re still not exactly sure what the best way to go about it is. We’ll look at some of the proposed methods in Part 3. First, let’s look at exactly why it’s important.
Data valuation is one of the best ways to introduce a data-driven culture across an organization. Data ceases to be such an esoteric resource once everyone within the organization can appreciate its value. Data valuation puts data on par with other valuable assets at a company, such as its real estate or hardware. With this, it’s no longer just the analysts and data scientists, or people crunching numbers in the finance department, who can appreciate data’s intangible value. Rather, a data-driven culture emerges in departments where it wouldn’t traditionally, such as marketing or HR. Having more individuals keen to work with data in their own operations, regardless of their role, fosters a data-driven culture which benefits the entire organization. Which leads us to 1.2: how data valuation contributes to better analytics across an organization.
Data is a resource, like wood, which you can apply to a beneficial end, like solving a problem or making a decision (or, to continue our wood analogy, building a house or burning for warmth). However, there’s some preparation required before you can extract value from data by analyzing it. Much like you need to varnish wood to build with or dry it out to use as fuel, you need to prepare data for analytics. This means grouping it into datasets, converting it to compatible file formats, and removing anomalies that will stunt your analysis. Only then does it become truly valuable. This is a significant side benefit of data valuation. It forces you to create analytics-ready datasets to increase the data’s overall value.
If you’re in possession of truly valuable data, you should invest in the appropriate levels of security. This could be cybersecurity to prevent malware or hackers stealing the data. Or it could be legal protection, to prevent third-party misuse or resale of your intellectual property. Neither of these data protection measures are cheap. Data valuation enables you to decide the level of protection needed have you take into account risk and ROI. For example, if you’re dealing with historical stock market data, there’s probably a relatively low cybersecurity risk because this information is available open source and therefore has little resale value.
Once you’ve created analytics-ready data, you’ve also, by default, created data that’s more readily distributable. You can distribute the data openly, as higher education institutions do with research. This brings value to innovation, policymakers, and public knowledge. Or you could distribute the data at a fee, which brings your company value in the form of a net-new revenue stream. Which brings us to a final reason why data valuation is important - data monetization.
Data monetization is the process of extracting commercial value from your internal data assets. To develop an effective monetization strategy, it’s vital that you understand the value of the data. As with any business, you need to balance cost of operations, marketing, people, with the value of the commodity you’re selling. The same applies to running a data business. If your data’s value doesn’t cover its expenses, your monetization strategy is no longer viable. Likewise, continuously valuing your data enables you to define to optimal price point to maximize your profit margins. So data valuation is one of the first steps to running any successful data monetization operation.
But what about the steps preceding valuation? Let’s look into them in Part 3: preparing for data valuation.
There are some preliminary steps which are essential for any effective and reliable data valuation. The first steps are concerned with getting your own house in order, and the final step involves looking outward onto the situations in which your data could be - and needs to be - applied in real-life.
Effective data management means that you’re:
Your data needs to be validated for it to be valuable. This means that you’re:
In this final step, you’ll answer questions like: is there need for your data? Can it be tied to a real-life use case? Which questions does it answer? What’s the market like for others selling this kind of data? Much like standard market research, conducting research around where there’s demand for your data helps you better understand your ICPs, the use cases your data is attached to, its VP as an intelligence solution. From there, you can market your data in the right places so that it’s received (and purchased) by people who truly see its value.
Moreover, market research gives you an understanding of the current market conditions in the commercial data space. As data commerce is still a young industry, this research can be more challenging than research, say, the market gap for a food or retail product. The best places to conduct market research for data are, funnily enough, on data marketplaces. There are over 60 in operation as of 2023. Use each data marketplace to understand the competition and price points for various data categories so that, when it comes to data valuation, you’re not over or under-valuing your product.
Researchers at the Harvard Data Science Review propose three different models for valuing data: the market-based model, economic model, and dimensional model.
The market-based model is unique in that it relies on income or cost to value data. Market-based models can be used to value data based on anticipated income, potential sale price, or estimated cost of a data breach or loss. Companies may also acquire competitors based on the estimated value of their data. It’s a model which ‘allow[s] for the monetary valuation of data based on what the market will pay, whether valuation is rooted in anticipated income, how much a data-oriented company might fetch in a sale, or speculation on what the loss of data is worth’ (HDSR).
The market-based model applies the same tried-and-tested method used by the IRS to determine the price of tangible assets and intangible assets like patents, copyrights, or software, but for data. The biggest drawback of the market-based model is that the data industry is still nascent, and so there are not always comparable products or companies to use as a benchmark for valuation. Likewise, data brokers did not tend to disclose their prices, although this is changing as data-as-a-service becomes mainstream and standard pricing is being adopted by DaaS companies.
The economic model values data based on overall economic and public benefits, which might consider job gains, privacy, health, and infrastructure. However, this approach can conflict with the market-based model. For instance, evidence-based health care relies on broad data from many sources, including providers, payers, and individuals, which an economic model would consider valuable for the public. Meanwhile, industry might prfioritize a market-based model to reduce costs or increase revenue for its sector.
There are to approaches to economic data evaluation. One is to estimate the value of open data. The second examines how policy can drive public data value. Both approaches to the economic model calculate the value of aggregating data from multiple sources. For example, the U.K. Hydrographic Office is able to generate £150 million annually (HM Treasury, 2018) after it made its data open and started sharing digital maps on surface water and geospatial measures. Now, the Royal Navy and 90% ships trading internationally use the UKHO’s maps to drive public value by delivering security and supplying goods. However, as a means of data valuation, the economic model takes a while to verify because, if miscalculated, the consequences can be significant across the economy, society and government. This is in contrast to the market-based approach, in which a private company can calculate the value of its data and can re-calculate it on the fly with little risk.
Numerous scholars attempt to value data via dimensions. These dimensions could refer to the data itself (e.g. its quality, recency, format) or the usage of the data (e.g. delivery frequency, user permissions, ROI based on the use case for the data). Various studies have proposed methods of calculating data value using these dimensions, with varying theories as to which dimension is most important when it comes to determining how valuable the data is.
The advantage of using the dimensional model is that it’s truly data industry-specific. It involves contextual attributes like data quality and stewardship, which other models omit. But the dimensional model invites contention between scholars, and is in some ways a more theoretical model. This means it could be less suitable for business application because you can’t quickly and simply calculate value - you first have to compare the different formulae on offer, then opt for the one which makes most sense to you and your data.
All three of these models have their pros and cons as methods for calculating data value. And each is only going to become more accurate as the importance of data valuaiton is acknowledge in data commerce, and beyond. So whichever you opt for, and whatever your primary goal of data valuation, expect developments in each field which will only benefit your valuation processes.
Having determined the value of your data, the possibility of data monetization opens up. For this, you need to think about starting a data-as-a-service company, creating data products, and finally, how to price these products.
Any company with exhaust data can launch a DaaS business. In theory, it’s a simple business model: you generate revenue from your company’s internal data. By monetizing proprietary data, your company helps meet growing global demand for external intelligence. You also add a net-new revenue stream using valuable exhaust data which would otherwise lie unused.
After data valuation comes packaging data into products and listing them on data commerce platforms, ready for purchase. The ‘product’ in question is usually a batch file of tabular data delivered into an S3 bucket. In this model, buyers purchase the data outright, often from a data broker or online marketplace. The data is then owned by the buyer and can be used however they see fit. The data is not hosted or managed by a third-party provider, and buyers must handle all the maintenance and updates themselves.
The price of your data needs to reflect its value and enable you to drive revenue. If you’re dealing in a certain data type, there’s probably an average ticket size for your products. For example, a location data subscription will usually sell for more than a B2B leads list. Nonetheless, you can create variation between your products by grouping them into one product, purchasable at a higher price-point. You can also adjust prices by charging more for more data maintenance, refreshes, and frequency of delivery. Lastly, the cost of your data varies depending on whether you’re selling a one-off data product, or a DaaS subscription. The cost of data-as-a-product is often higher than DaaS, as buyers must pay for all associated costs upfront, as opposed to the subscription model usually attached to DaaS.
As we’ve seen, data valuation is an essential process for any organization that wants to better understand the value of their data. By determining the monetary value of datasets, organizations can prioritize data-related projects, make better decisions about how to handle their data, and cultivate a data-oriented mindset across all departments. Additionally, data valuation contributes to better analytics, data protection, data distribution, and data monetization. While there is no one established method for calculating data's value, understanding the proposed methods can help organizations make informed decisions, and add on-top monetary value to an organization.
The easiest way to take action after valuing your data assets and begin monetizing your data is by joining Data Commerce CloudT™ (DCC). DCC is a SaaS product designed by Datarade to help companies sell more data. It’s the easiest way for data providers to build an omni-channel data business. With one DCC account, you can publish data products in your selected data marketplaces and channels - instantly. To learn more, talk to our partnerships team.
List your data in the most popular marketplaces.Learn More