As most industries wake up to the digital age, data deluge becomes a natural consequence. How does one manage it effectively? What are the key elements to make data productive and realise its value?
A life insurance CMO I met recently told me that they have done away with physical forms. Every application and document has to be submitted in an electronic format. No exceptions.
So, that is rule # 1 — discipline your data capture mechanism. If you are a legacy industry, invest some effort in data capture/enrichment.
You have agencies that can do this for you on a per record basis. This will save you serious dirty labour down the line. A PSU bank did something similar years ago before it embarked on a larger data integration project.
Rule # 2 — Jot down how you will make use of the data. Who all will be involved, which other departments need to be co-ordinated, who will close the loop. You don’t have to be 100 per cent certain, but should have some sense about it.
Almost four-five years ago, one of the Top-3 private sector banks adopted the practice of having a dedicated department servicing the data/analytical needs of all other departments, called a BIU — short for Business Intelligence Unit. This ensured that the department making the request was thinking through what it would “do” with the output.
The processing department, over time, knew how the data was put to use across the organisation and what were the emerging data/analytical needs. The BIU also became the driver for putting in the data discipline mentioned in rule # 1 above.
Next — start with data integration. Integration comes from the Latin word integer , meaning whole or entire. It has been defined as the bringing of people of different racial or ethnic groups into unrestricted and equal association, as in society or an organisation. Just replace people with data and society/organisation with store or base.
The idea is to develop a common base from which you can make sense of the data. So that the analysis is reliable, trust-able. It is not based on some person’s laptop data, or another one’s individual definition of accounts receivable. Start small, may be with two-three key functions or outcomes, never with source systems.
The last step is realising the value of the above three rules through analysis — What happened? Why did it happen? And, what will happen? Or as I read a consulting company put it — “hindsight” and “foresight”.
The first part is answered by the standard reports you get from your enterprise resource planning (ERP)/operational systems — monthly production/consumption, 30-days outstanding, inventory position at each depot. The second one requires you to adopt a BI (business intelligence) tool that will help you seamlessly manoeuvre the entities/data/reports/key performance indicators (KPIs), etc. Compare sales with same period last year, across territories/SKUs, what-if I changed my plan by 5%, increased ad spend by 2% and the likes.
The last one — what will happen — is a science in itself and deserves a column all by itself. Simply put, business and technology come together with statistical skills to predict the future outcomes — which customers are likely to buy this new product? Which distributors are likely to reduce their business with me? Which stores are more price-sensitive than others?
Get these four parts together and your journey of making data actionable will start yielding results … and money.