How Much is this Data Worth?
Complex data analytics are more realistic than ever before thanks to great improvements in the ability to capture and connect more data to the physical world. One of the most ambitious concepts currently circulating this network is that of the “digital twin”. The digital twin concept proposes a virtual clone of real life things such as buildings, infrastructure, products or processes, then replicating how they would behave in the real world. If that were achieved, opportunities could open up to analyse real time performance and test “what-if” simulation scenarios in the comfort of the virtual world.
As the demand for capturing, consuming and analyzing data rises, how do we ensure that the datasets actually produce the value that we hoped for? To answer this question, this article explores the potential benefits of valuing datasets in monetary terms.
Money, Money, Money
It may be counterintuitive to focus solely on the monetary value of data when, as professionals, we are much more comfortable talking about worth in respect to the data’s technical merits. However, when it comes to discussing the worth of data with others outside the industry, this message is a little trickier to convey.
One option would be to employ a metric or grading system to help others understand our data and its value, but these inevitably require translation and context. For example, what we might rank as an “A” grading might not translate to someone who uses a system based on another metric, such as stars. Money is a language that everyone understands and provides a simple consistent value that is easy to make comparisons. Notwithstanding, it can widen the interest of certain data to other groups such as boardrooms, or external investors.
How Much is this Data Worth?
You may be wondering how is it possible to monetarily value data? While not an exact science, the starting point is to apply understanding of how each data record will generate revenue or savings over its lifetime while considering the underlying costs.
A data record could generate worth a number of different ways but some examples include:
- If the data enables an operational savings by removing the need to carry out a site visit, such as to measure a height clearance underneath a road bridge, the record would generate value by removing the need to disrupt road users.
- If the data is of sufficient commercial value then it could earn revenue through a data subscription service or licensing agreement, allowing others to view and use the data. If there are significant barriers to entry for others to collect or source a similar dataset this option might be of particular interest.
Regardless of how useful the data is, there is no such thing as free data. From its birth, data accrues costs to cover its creation. For the rest of its lifetime it will incur costs to host, secure and administer the data.
As the data ages (and especially if the data relies on human input), it will typically become outdated or incomplete, and the cost to maintain data integrity will need to be factored. If the integrity of data deteriorates then the data will become worthless at best, but in the worst-case scenario, it will become a liability. For example, if someone were using outdated height clearance data to determine if they can pass a tall load under a bridge, and there then happens to be a collision between the load and bridge, it is safe to assume the data is a liability with negative worth.
What are the Opportunities Once I Know the Data’s Value?
Once we have a snapshot of the data’s value what can we do with it?
If you have a valuable dataset then options for protecting its value can be explored. This can be through enhanced security, improved infrastructure or even insurance. In terms of protecting data integrity, we can also use a positive valuation to justify investing in maintenance spending, perhaps through better training or systems that are designed to minimize the impact of human error.
In cases where the costs of an existing dataset outweigh the benefits, opportunities arise to create solutions that reinvigorate its worth. For example, if you are trying to maintain a very detailed dataset requiring lots of human input then we could suggest rationalising the number of fields we are updating by valuing each of them to reduce the burden. Secondly, we could justify investing in new technology to minimize human input and make collecting that information more cost effective and reliable. Another solution for a low value dataset is to advise on decommissioning the data to free up resources.
If we are at the stage of creating a new dataset, by forecasting a dataset’s value at its design stage it can form part of its business case and help to highlight the future benefits and costs (such as maintaining data integrity) of the data for securing a future budget.
In conclusion, with increasing desire to create new digital data that simulates more and more of the physical world, we can use the insight provided by a dataset’s monetary net worth to rationalize decisions about virtual data’s purpose, design and future.
Chris is a data analyst in the IBI Newark Office (UK) with 10 years experience in data analysis and consultancy working in the transportation sector. Chris is currently collaborating with colleagues across the UK intelligence sector to integrate Highways England’s multiple primary inventory datasets into IBI’s AVIS (ASSIT) platform, and currently heading up the technical delivery of the Mast Review Project which involves using the same platform to locate and capture previously unaccounted for structural masts and large signs on the UK road network.