Being data-driven is being context-driven

by bold-lichterman

This is the big cream pie of the moment. All companies proclaim their desire to become “data driven”, in other words to be data driven. Is this such a deep break with the past and why is there so much emphasis on this goal?

Companies have always been data-driven

As surprising as it may seem to you, we have always driven companies by data, or in any case by numbers, mainly according to performance and profitability indicators. Believing or making people believe that for more than a century one has made decisions on instinct and with a wet finger is undoubtedly an easy way of selling new approaches. Ultimately, companies were too often criticized for looking only at these indicators and therefore for making biased decisions because they were not sufficiently informed. So, no, data is absolutely not something new and what we present as a revolution today could have been seen as a flaw yesterday.

When the raw indicator is enriched by understanding the context

This is where we talk about Big Data again. Finally, note that for a year we have been saying more data than big data because, here again, in terms of data we have always done as big as possible. When you remember that in the mid-1980s a PC with 128kb of RAM and a 5 ″ 1/4 floppy drive was a powerful machine and that today the simplest of smartphones is infinitely more powerful than the onboard computer of the Apollo capsule that took Man to the moon, we can clearly see the progress made in terms of storage and processing capacity and that if yesterday’s big has nothing to do with that of today he was still the big one of the moment.

The current storage and processing capacities allow us two new things. The first is, precisely, to take into account the sources and quantities of data that are infinitely more varied and and in greater quantity than before. The second is to process this data to identify correlations, “patterns”, between very diverse data.

Ultimately, this allows us to look up from the final gross indicator to take an interest in its context and what impacts it.

Some examples

In terms of monitoring commercial performance, we had the numbers and the pipeline. Sometimes in connection with the marketing effort but as these two functions still speak too rarely to each other, it was already complicated to think of managing the two together. Today we can take into account a wealth of data relating to, for example, the climate, exchange rates, HR factors specific to salespeople and an infinite number of things depending on their relevance to a given sector and market. . Ultimately, this allows you to make decisions in terms of production, pricing, promotions, etc. based on weak signals sometimes located light years away from your business.

Ditto in HR to anticipate the risk of attrition. The risk of seeing an employee leave can sometimes depend on measured and identifiable managerial or HR elements, sometimes totally external contextual elements. For example, it has been shown that one of the motivations of a departing employee is a certain number of departures from his relatives and that the notion of loved one was not necessarily estimated at the level of the members of his team but of people with whom he had the most frequent interactions on the corporate social network.

In short, the interest of data is not so much in the key indicator which is often simple and known but in the understanding of the context, exogenous factors sometimes very distant from the business which condition this indicator. A bit like thebutterfly Effect.

The value of data is not to improve existing indicators but to provide others that help to understand them.

This may seem obvious, but we still have to draw the consequences. It will often be necessary to seek data in a perimeter outside the subject concerned, sometimes outside the company or even outside the business. And therefore be inspired because we cannot deal with everything: we must therefore seek and judiciously select what we want to deal with and improve the system empirically.

Data is not the end of judgment

That said, we must not believe either that switching to a “data-driven” logic will spell the end of judgment and individual appreciation. As this McKinsey article clearly shows, data and humans are complementary. As powerful as it is, data only gives a fragmented understanding of the context. What the data does not give is the story of a human context. (“[What] I don’t think the data did do, was [have] an understanding of human story and human narrative. The data isn’t going to produce that for you ”.). The solution is neither in human expertise nor in the data but in the way in which we manage to articulate the two.

In a way, I would be tempted to say that certain disciplines such as marketing or even HR will, in a world driven by data, become more of the arts than sciences or rational expertise.

At this stage a question still crosses my mind. When a company proclaims its desire to become data-driven I interpret it (rightly or wrongly) as:

  • a desire for “modern” display using the buzzword of the moment?
  • a desire to reassure the market and investors on the business orientation of digital projects alongside initiatives stamped “experience” whose qualitative and non-process dimension does not always convince addicts of ROI.
  • a desire to remove any bias from business decisions (even if it means neglecting common sense) or in any case to send a message in this direction to reassure.

So I hear a lot about deciding, not understanding. Bad wording or tendency to neglect an essential part of the contribution of data to the management of the company?

The expert:

bertrand-duperrinBertrand Duperrin is Digital Transformation Practice Leader in Emakina. He was previously consulting director at Nextmodernity, a firm in the field of business transformation and management through social business and the use of social technologies.

He regularly deals with social media news on his blog.