BETA
This is a BETA experience. You may opt-out by clicking here

More From Forbes

Edit Story

Data Science: Getting Real

This article is more than 9 years old.

Consider This: Which would you rather have at your company? The world’s greatest data scientist working alone in a corner lab or data that will make all of the employees of your company one percent more productive?

I suspect that most organizations, with a little thought on the subject, would choose the latter. And yet, as companies enter into the world of Big Data most seem to be choosing the former. Indeed, ‘data scientist’ is one of the hottest new corporate jobs around.

The goal of this column is not to diminish the profession, importance or role of ‘data scientists’ throughout our industries. Most of the important innovations made in Big Data and corporate information have been, and will continue to be made, by data scientists. In the years to come some great data scientists will undoubtedly change the way we see and understand the world. This is all greatness that we will benefit from but one fact remains: there are not enough data scientists to go around.

As always the purpose of this column is to separate the practical from the hype and to establish real-life priorities that will ultimately benefit your enterprise. And so, I would encourage each of us to pause and ask ourselves one fundamental question: Does your organization really need a data scientist?

I believe the answer for most organizations, at least for now, is No.

Why? It’s not because data scientists aren’t valuable. Rather, it’s because most companies have more pragmatic business priorities regarding their data; initiatives that take precedence over specialists singularly focused on making esoteric breakthroughs. For most companies the critical part of R&D is the ‘D’ of development; so too for many companies, the emphasis in the DS of data science and the focus for now and the near future should be the ‘D’ of data. And while it’s important to conduct research science it is also likely that operationalizing data will readily and quantitatively contribute to the bottom line – that is, getting the right data to the right people in the organization in a form that they can immediately put to use.

The reality is this: most of what data scientists do is essentially science. One and done, often with little repeatability, scalability or practical application to the enterprise. Moreover, most of the analytics they perform target only a small set of people or fairly narrow, focused problems. These specific problems may be very valuable to organizations especially ones that can automate certain functions algorithmically.

There is nothing wrong with that. On the contrary, it was the predecessors of data scientists working in great labs at places like AT&T, Bell Labs, Xerox PARC, DARPA and a host of universities that gave us every major step of the electronics revolution from the transistor to the Internet to Big Data itself. The value of these brilliant individuals to the modern economy is vast and undeniable – but those breakthroughs were also hugely expensive and took years, and uncounted false starts and blind alleys, to achieve. Being a high risk, cutting edge activity most of this work also ends up shelved because it doesn’t have a lot of actual applicability. The real question you should be asking is whether that is where the data can provide the greatest value to your organization.

I suspect that most of you reading this, unless you are working for one of the world’s largest corporations, it is probably not. Rather, you want to keep your products competitive, serve your customers better and out-distance your competitors by creating, uncovering and unleashing competitive advantage. For that you don’t need the rare data science breakthrough but a constantly improving ability to access and put to use your growing mountains of information.

More and more companies are focusing on the data scientist. There are a lot of companies adopting the latest tech fad, without really understanding how to use it, because shareholders and customers like it – and because their competitors have already done so. It is no different with data scientists. By hiring a data scientist, CEOs and the executive staff can shift their focus away from their growing data issues and defer what is important about all of that new data to someone else that is deemed the “expert,” regardless of whether they are handing it to a person who may or may not be particularly business savvy or have a broad connection to the core business.

This approach opens a competitive opportunity for smart companies. Those companies – and yours can be one of them – aren’t depending upon an expert but rather are thinking about how their organizations can work with that expert to then use the data that is being collected and curated and ultimately find the value in that data for every employee.

What those companies are coming to understand is that Data is the fourth strategic corporate asset.

A little history. The first strategic asset, identified in the ‘70s and ‘80s, was the use of hardware and software to get a handle on all corporate financial and operating capital assets. Think ERP (Enterprise Resource Planning). For the first time companies could understand how they were performing financially.

The second strategic asset, identified in the ‘90s, was customers. Think CRM (Customers Resource Management). Now, again for the first time, companies began to look from a holistic, customer perspective. It proved to be a major asset improving product development, sales and most of all, customer development and retention.

The new century also saw a new wave – this one less defined: the management of employees as strategic assets. Think Human Capital Management. It is a revolution that is still underway with innovative breakthroughs occurring everywhere from recruiting, to rewards, to the shift to part-time contractors and virtual employees.

It is now apparent – and I would propose - that the next, fourth, strategic corporate asset is data. And just as with the first three, none of which was implemented by some expert going off to a corner in search of a breakthrough but rather the benefits were gained by systematically managing the entire enterprise every day. The same thing now needs to be done with data. It needs to be managed strategically as a core business operation not a new wing in R&D. It also must be operational, impacting every employee in an organization, allowing them to be more effective in their roles.

The successful companies of the future will be those who have learned how to operationalize their data, not silo it – and to approach that data from a human, employee perspective, rather than from an experimental, scientific one.

In the next few columns, I’ll describe how that can be done . . .