What makes data valuable?

This week’s blog comes from Anna Faelten, an Associate Partner in our TAS TMT team

Data, data everywhere…but what actually has value? Do you have a hidden gem?

The value question is one that I am increasingly being asked as the digital age allows – and indeed encourages – organisations to collect increasing volumes of data and apply progressively sophisticated analytic techniques.

But the fact is that the more data we collect – and, in particular, the easier it becomes to collect – the less valuable it becomes. An excess of data can indeed be dangerous if collecting it becomes an end in itself.

So, in this blog I’m going to examine the qualities that enable data to command a premium.

The value paradox

Data has always had value.  The Babylonians carried out a census to determine how much food they needed to feed their population. The ancient Egyptians used surveys to assess their workforce. The Normans infamously undertook the Doomsday survey to better understand their ‘acquisition’ of England in 1066.

What’s changed over time is how advances in materials, and more recently technologies, have changed our ability to record, store, analyse and share data. Indeed, the ability to store and quickly analyse vast amounts of data is one of the major differentiating points of this digital age. Data has become the basis of competition in many sectors.

According to McKinsey, 90% of all data in the world today has been created in just the past two years.

But, what hasn’t changed is the fact that not all data has equal value. In the examples given above, the agencies collecting the data had a clear purpose – to create useful intelligence. The data collected helped the agencies to improve their understanding of food demand, collection, distribution, and associated operations – and to allow for improvements to be identified and implemented. The intelligence was highly valuable by virtue of providing unique, clear and actionable understanding.

This stands in great contrast to the vast amount of data available today. A lot of data collected today emanates from social media, or online interactions, and is only useful when analysed in large quantities to identify trends and sentiments. The digital age also makes data very easy to collect, store and distribute – which at least puts it in danger of becoming commoditised.

This begs the question: what data still has value?

Lessons from data’s front line

Transaction data on the B2B Information Services holds some of the answer. The charts below show the average and maximum transaction multiples paid over time within the sector over the last ten years. This is during the period of greatest data expansion and arguably commoditisation and yet, for this dataset the average EV / EBITDA multiples have risen from c. 10x to c. 15x and maximum multiples have risen from c. 13x to c. 26x.

B2B Info Services Transactions

 

Even more pertinently for our analysis, the spread between the maximum and average multiples has risen from c. 3x in H1 2008 to c. 11x in H2 2017, with buyers paying an increasingly high premium for the highest quality assets.

So what are these premium assets offering? If we dive deeper, we can see that it can be broadly categorised in one of three areas

  1. Proprietary data that requires physical capture and/or specialised access
    There is still data that needs to be collected by humans and data that comes from sources where one or a limited number of organisations or individuals have limited and specialised access. The need for human legwork, physical infrastructure and unique access intrinsically gives data more value. Examples include the data collected by journalists from their personal sources, the data collected by cameras from ports or data from a network of sensors


  2. Data accumulated through long-term aggregation
    Data invariably disappear over time through design or through accident.  Longevity would tend to indicate a high level of interest and therefore value. But, more than this, the time taken to accumulate the dataset can also add data’s value if it isn’t easily replicable. This inherently makes defendable and therefore more valuable.


  3. Imbedded data
    Data that is collected and shared with customers through embedded workflows will also have more inherent value since the very act of data collection depends on a relationship, built on trust, that isn’t easily replicable and again is defendable. The data collected is also highly relevant and attractive to the customer. This can also include revenue streams linked to IoT-derived data, collected on behalf of engineering companies and used to schedule MRO (maintenance, repair and operation) services.


From this understanding, we’ve developed three simple questions that help us to assess the value of individual datasets, looking across the attractiveness of the underlying market, the value to customer and the defensive qualities of data in that subsector.

How attractive is data

To determine the value of their data to external parties, organisations should speak to their potential customers and consider the wider market. Have others tried and failed to find the hidden value? It’s likely that unless your organisation has unique access or scale, it will be difficult to derive significant value if no-one else has done this before.

Big data

In this context, I’d like to put forward a word of caution on ‘big data’. As mentioned above, data has become the life blood of organisations. The term ‘Big data’ refers to the huge and increasing volume of the data available, and the ways it can be processed. But the value of big data lies in the actionable insights that your business can draw from it, rather than the size of your dataset. If companies collect the wrong type of data or ask the wrong type of questions, the process could be distracting and ultimately harmful.

As my colleague noted here a couple of weeks ago , the sheer amount and complexity of data available to companies means that they’re not always looking in the right places or getting the most out of their data. Essentially they have too much information and limited insight. Companies that crack the difficult task of to systematically collect, clean, and store data in an actionable way which makes it insightful to clients will be the real winners