Calculating the value of data is something I have been thinking about a lot. Data, any and all, seems to be relentlessly hoovered up whenever we use any form of connected device. Who had ever heard of data as a mainstream topic twenty years ago? Nowadays, we have seen Mark Zuckerberg answering to Congress in the States and countless articles based around what Google and Apple know about us. Some people are laissez-faire about it whilst others veer towards the downright paranoid.
Organisations collect data, they hoard data and (hopefully) guard these vast amounts of data that they collect. Why? Because it is valuable. It is useful. Apparently. However, who in a company actually gets down to the nitty-gritty of this and can measure and express the Return on Data that this feverish collection and hoarding actually brings to the organisation?
In 2015 Doug Laney from Gartner wrote about data in financial terms. How it can affect the value of a takeover target, if they have a vast unexploited data store, for example. Were that to be monetised then what is it worth? Does this mean the buyer is getting a fantastic deal or when it seems to be overvalued on traditional metrics is the difference made up by the value of their data? Herein lies a real problem as the difficulty in valuing data stems from several reasons.
Firstly, that there is no firm formula to do so, because to some that data is just wasted storage and to others it is gold. Whereas a physical asset, such as a piece of land, is such a mainstream asset that it is far easier to value. With data, the great big lump of bits and bytes only has value if the owner knows how to extract information and insight from it, and use that effectively to make them more competitive or to sell to someone else in a finished and usable form. People have had a stab at it by trying to vary old maths and make it new maths. I found the following on the Internet:
Though his looks like an elegant formula, the Gain from Data metric is subject to so many other variables, primarily time, it is almost impossible to calculate so simply, makinf the formula impossible to scale. It only serves to highlight just how the temporal aspect of data value is so important. Depending on what it is, it may be very time limited, making it useful only in a very brief window. Think of data like a paper currency that can burst into flames at any moment.
One second it has the face value and the next it is ashes.
In contrast, a piece of land is just there. No more land is being created, whereas data creation is never-ending: limited only by our ability to get it and store it.
Secondly, the technical aspects are crucial. What form is it held in, on what type of database, where is it held (there are massive regulatory differences around the world), have the data owners consented to its use, by whom, how old is it, how consistent is it and so on. If I can’t use in my company for my purposes then it is just Ones and Zeros on a hard drive somewhere, merely cluttering up the ether. Utterly without value.
The fact remains that extraordinary amounts of data are being recorded about us, all of the time. I recently holidayed in Norway and in ten days I didn’t use one bit of hard currency. All card, all the time. I navigated around using Google Maps. I checked TripAdvisor and used Uber, as well as uploading countless photos to Facebook for family abroad to see. In doing so I must have left an enormous digital smear across the Norwegian landscape. Me and the thousands of other tourists on holiday at the same time. Can you imagine the quantity of data generated by me and the billions of other people using connected services every single day?
To be able to achieve a RoD that makes all the efforts and costs at collection and storage worthwhile, several things need to happen and I can only really see that these can happen under the guidance and direction of a very senior – if not on the Board type senior – individual who guides a team with specific responsibilities. Call them a Chief Data Officer (CDO).
Ideally the value of data is considered so important that the CDO is on the Board. The CDO would need to have close ties with Marketing and Strategy functions to understand how they intend to use resources to achieve them, and whether existing data is useful or new data needs to be acquired. Additionally, they need to know how to shape and deliver it to them in a worthwhile manner. Then there needs to be a real-time feedback loop – Sales? – in order to assess the efficacy of the deployed data as well, as a direct line between them and the technical functions of the company. The sort of things CIO deals with, especially storage and access. The CFO will have demands on their funds from the CDO. They need to be able to understand the RoD and how it is affecting the bottom line, the share price, their partners and so on.
Most importantly of all is someone who can see through the Fog of Promise that all this data is purported to hold. The RoD that can be achieved if only they used it ‘properly’ is the sort of golden thread that is so often sold to them. Correlation does not equal causation. I’ll repeat that: correlation DOES NOT equal causation. Falling into the Feynman Trap is something that affects the best and the brightest (Famously, Jim Collins did this in Good to Great). Usually when they become mesmerised by their own belief in the infallibility of data.
The CDO not only ensures the data is valued correctly, they are responsible for preventing their company being led down a rabbit-hole of promise of the jam-tomorrow variety. The sunken cost fallacy remains as relevant today as it ever was and sometimes the emperor is indeed naked.