Have you wondered in management teams or by business decision makers what are the biggest real costs of data practices over a number of years? In many cases, this mindset impacts the pricing, profitability, and investment decisions of data-driven businesses and services, but in many cases, this work is not considered long enough.
Let's take an example. If a company is launching an Internet-of-Things (IoT) implementation, that is, installing sensors in a semi-large environment, collecting information about them every hour, combining the collected data with data from three other information systems, and storing everything for at least ten years. This data is visualized and analyzed for your own use, so the business value of the data is weighed against the customer and business value brought by that analytics. Unfortunately, many times this is done by purchasing that analytics service or application and the investment decision is made based on the offers received from potential vendors but in below we will create an additional point of view for this decision.
The prices of analytics or artificial intelligence services are evaluated first. Let's just say that the artificial intelligence service that suits you costs 1,500 € a month and the contract is made for three years, which means that the cost of 54,000 € is entered in the investment decision. Let's imagine that storing one data packet with metadata and security costs 0,01 € / month / sensor. If there are 10,000 sensors in the company's and customers' environments. As a result, over a ten-year period, the cost of storing the data is 0,01 € * 10,000 sensors * 12 months * 10 years = 12,000 €. This amount does not include maintenance fees, possible security upgrades, additional developments, or other things done during the lifecycle. If this data is combined with data from three other systems and the cost of one integration is estimated at 15,000 €, this means 45,000 € in deployment costs. Once again without lifecycle maintenance fees.
If business leaders do not have the tools or expertise to estimate the true cost of business-driven data generation, KPIs for data-driven services may appear with red flag throughout their lifecycle. In the example given, the cost of an artificial intelligence system more than doubled over its life cycle. Think about how many systems you have, where to get the data, in what lifecycle, and how many tens of thousands of sensors are to be deployed in total?
These figures and calculations in the example above are inaccurate but still sufficiently understandable for each of us to really start to think about what the real costs of data practices are and how for example Return on Investment (ROI) or breakpoint should be calculated.