Balance sheets are funny things. In one stroke, they tell your positives and negatives. And it takes no time for assets to convert into liabilities also. The best examples of assets and liabilities can be taken from our daily life. If the kid turn out to be good (whatever it means to one), the kids are assets otherwise they become liability. A marriage or a partnership can become both asset or liability.
Companies should also look into their data from a similar perspective. Data in itself has no meaning if you can't mint it. You can boast of having Terabytes and Petabytes of data, having clusters and farms and ranches and villas and islands of servers to store them. However if you are not able to make any money about of it, is it even worth the salt?
Ask a plain question about your data. Is your data eating money or earning money, net-net? If you are data is eating money, and you don't see any concept of breakeven or may the concept of breakeven does not exist in your organization, than you are fishing in trouble waters. Data seems to be a very romantic thing. Don't get me wrong, but I have seen people getting in love with their data and protecting it for ages. And it's not a new found indulgence of humans. People have been collecting stuff throughout history, which in some form is a data. However at personal level we are talking about a cupboard of data. Also the speed with which the data was getting generated was also humane. But not with the so called "Big Data". The speed with which data can be generated can quickly become humongous. Write a while loop with true as condition, and soon you will be the proud owner of Big Data. Feeling proud of it.
Let's come back to our original line of discussion. Is the Big data an asset or liability? It's important for organizations to have a clear insight into the cost of acquiring data and the money they are making out of it. Develop the unit economics of data. Develop a notion of unit of data in your organization and see what it costs to acquire and maintain it. Also see what kind of money you are making out of it.
Let's take a simple example from IoT domain backed by sensor data. From the cost side metrics look for the following:
- Landed cost of sensor to your customer premise
- Cost of installation
- Cost of storing the data in the cloud per sensor per year basis. (I know you love to put your data in cloud)
Now look at the revenue side of equation. Let's say you have sold it on per sq. ft. basis. That doesn't matters. What matters is money you are getting per sensor basis. Sensor is your unit of work and internally from costing perspective that's what you should be factoring in. So now it's a simple equation:
Let's say you have got N sensors in a premise:
a = Landed cost + installation cost
b = cost of storing data per year basis. (I am ignoring the time factor as usually the pricing is done on per year basis)
So Total Cost = N(a+b)
Let's say your revenue is R so per unit Revenue R/N.
For operating breakeven
R/N = (a+b)
=> R = N(a+b)
This basically means that with N sensors the revenue that you should be looking for. This is operating breakeven just for the sensor part of the things. We are still far away from the True Profit.
So have a very clear idea of your a and b for sensors to understand the Sensor's economic model and which in turn can be applied to Big Data economic model. a is the cost of putting the system in place which will capture the data per unit basis and b is the cost of capturing and maintaining the data.