What is REALLY “the Value of Data” …?

Everybody talks about “data as a key asset”. Everybody knows that data has value. The Fortune 500 companies spend billions in data management and governance. And yet, having browsed through dozens of their balance sheet statements, I have still to find a “Data” line in it.

Take Google Inc. for instance. Google is all about data. If there is one company in the world knowing the value of the data asset, it should be Google. But if you are looking for “data” into Google’s balance sheet, you’ll find… nothing, not even within the “Intangible asset” line. Google does not seem to “monetize” its data asset.

What is the value of data, then? 

Most companies are using data to support internal purposes, such as running business processes via system transactions, produce reports and generate decisional information. For them, the equation seems simple: no data, no business. It would be no stretch to claim something like:

Value of data = value of business

Now, wait a minute. We do live in a highly digitalized world but there is still some physical reality remaining down here. That means: if we delete all product data in our lovely ERP systems, the goods would still exist in the physical inventories, no matter what the system says.  

This is because data is nothing more than a representation of real-world objects (products, customers, suppliers, assets, etc.), like a map represents the territory. Here is a truth that derivates from it:

Data is an abstraction of reality.

Losing product data would prevent real-world goods to be sold and moved out of the inventories, and impact the Profit & Loss statement. The reason is that data is primarily used to run operations in the information systems, which enable the execution of the business processes. 

Data is like the map of an unfamiliar location in a GPS navigation devices: we may still reach the destination, but at the cost of a much less efficient itinerary, wasting time and resources, and maybe losing opportunities.   

Besides, all data is not equal. Losing the map of a location we don't intend to visit does not have any consequences. Similary, the fate of obsolete or inactive data would not impact business activities. It could even go unnoticed forever.

But if we change our travel intentions, the situation changes: a data record that has a high importance today may be completely irrelevant tomorrow, or the opposite way around.

This means that data has a dynamic value that is dependent on its actual usage. Advocates of the hard valuation of the “data asset” may start to worry, but the real shock is still to come...

Value itself is a fluid notion. For a salesperson, the value of a product is conditioned by the order to cash process, which starts at sales quotation and ends up when the invoice is emitted. For marketing people, the value of the same product is expressed by the total sales potential over a certain period of time. For an inventory manager, it would be the insurance cost of the products in the warehouse. Three perspectives, three ways to measure the value, but… it's still the same data object.

If we want to value product data, how would we define value in the first place?Let’s face it, there is no single, universal truth as each business perspective is correct in its own context. Giving data a unique value would be, by definition, an arbitrary exercise that would let the door open to criticism and mistrust.

Now, we need to consider why we want to give a value to the data asset for in the first place.

The intention is good: data is too important to be considered as “cheap” or “free”. Valuating data would compel everybody to give it the appropriate care and attention, and secure quality on the long term.

There is no question: we need to reveal the value of data, in a credible manner that emphasizes the need for quality data. 

Data is an enabler for the business processes. The finality of the business processes is to enable the generation of value, e.g. sales, benefits, cost avoidance, etc.  Therefore, data enables value generation.

But data enables value only if it is of the appropriate quality. This translated into simple truths: 

Good quality data enables realization of value

 Bad quality data brings risks to the realization of value

Quality is rather binary: data is either good (enabling value) or bad (brining risks to value), which implies that

Total value = value enabled by good quality data + value endangered by poor quality data

Here, we have a nice “equation” in which data contributes directly to the business value chain without having to carry any intrinsic value. This definition is all but innocent as it means very tangible things:

• Only data that is in use within a value chain can be assigned a value;
• The value assigned to a specific data set depends on its contribution to the value chain at a specific time: data contributing to big business will naturallly have a higher assigned value;
• Poor data quality affects directly the value chain in a measurable way, that determines the priorities of resolution (big impact first);
• As the notion of value may depend of different perspectives, the same data can carry different values for different value chains.

By adopting this definition, we introduced a dynamic measurement of data that can be used to demonstrate the level of contribution of proper data quality to the business and highlight the business risks of poor data quality.

All this seems to be a nice theory, but it has become a reality for few companies at low effort. It may be surprising, but measuring this dynamic value is a straightforward process for organizations that run business processes and have some data quality management, even of relatively low maturity level. It's very often a low-hanging fruit.

This is because the determination of this dynamic value is mainly a matter of approaching the problematic from a different angle. In most cases, it just requires to contextualize the data quality systems. It's neither difficult, hard or painful: in my experience, the most difficult is to change the mindset.

How to do it, practically? I'm sorry for the cliffhanger, but this will be the subject of another post!

Stay tuned.