It open concept verification to achieve large data win

This approach sounds reasonable, and the study of leadership change will propose a different approach. Although proof-of-concept seems logical, value validation is more pragmatic.

When I browsed the most outstanding literature from legendary writers Kurt Lewin and John P. Kotter about leadership change, I did not find the concept of "proof-of-concept" anywhere. In fact, I stumbled across the idea that the only place was in the IT community. This is because the IT staff are from the engineering subculture, and in their culture you cannot publish anything that is not perfect. Proof-of-concept is a method of using experimental methods to reduce risk without releasing passable products.

However, questions about concepts-even if they have been proven to be inherently deficient in their commercial value. On the contrary, what I call value validation is more commonly called small success with an astonishing amount of business value. Since it involves one of the biggest challenges in working with large data, let's compare the two approaches: executive sponsorship and leadership.

What are we going to prove?

Executive support is one of the biggest challenges in today's relationship to any big data work. Executive directors understand that because of all the inevitable hype there are some things; however, you cannot assemble powerful execution support until you have a strong business case.

When I help visa for its corporate data strategy, we spend weeks gathering strong business cases because it's the only way to get executive attention-and more importantly their money. Big data work hard to compete for executive support, because from an executive's standpoint it looks like it's fooling around without any real business purpose.

True executive leadership often appears too late in the proof-of-concept model. The CIO will find some money to buy expensive technology and work with its highly-paid data scientists and consulting team to start data mining.

That means combing the company's data to find out what it is that no one knows. The general manager of the business unit waits patiently for it to discover something useful. If it actually manages to prove a concept, the general manager will perk up and the big data job may run away. This is when executive leadership and sponsorship start-if it does happen.

This approach has a lot of risk. The actual currency is spent, but there is no indication that these analyses will have any commercial value. If the CIO spends too much time (the timetable is generally not shared with the CIO), the patience of the senior management will vanish abruptly, and the whole effort will be shelved until the next wave of analytics hype is over.

Lead a better model

To ensure the real value of your large data strategy, you should boot correctly. The value validation model is either driven by the general manager or by the CEO, depending on the scope of the work. The appropriate vision established in the front is to rely on commercial value, and then large data is transferred to support the work.

Following that vision is a small success, and a greater triumph of reliance on early momentum. The final big data strategy solidified into the organization, becoming what Kurt Lewin called a new freeze. This is the leadership change that every executive director is familiar with.

Having a CEO or general Manager to guide a large data strategy in this way completely eliminates the risk of inadequate executive sponsorship, so it's one of the biggest risks I've seen in today's big data jobs. It also encourages the success of the leaders who are more likely to return the output of your large data investment.

As I stated earlier-this is culture. IT managers, and even CIOs, pretend to be overly secure in their proof-of-concept mindset. Proving value has more business significance and accelerates your common understanding of the value of the assumptions. You learn more from the pursuit of small success failures than from successful proof-of-concept.

Most importantly, the method will always hire executive sponsorship as long as it is meaningful. In large data experiments, executive managers have abandoned a lot of money and time because they have lost confidence in them. They lose confidence because of inadequate participation, which leads to miscommunication around opportunities.

This is the key to failure with the concept validation method. In contrast, when business executives are committed to this, they closely monitor the opportunity. Regardless of the outcome, small successes lead them along the company's most meaningful path.

To succeed in your big data strategy, you will need an execution mindset-not an engineering mindset. It's understandable that engineers spend a lot of time doing things. We need engineers to be accurate, but in a place where values work, this mindset doesn't fit well with business strategy.

If your company is currently experiencing big data trying to prove a concept immediately-interrupts. You are wasting money. Instead, tell your general manager to create a close business case and reposition the direction of your large data team. If you intend to prove something, it will become more valuable.

(Responsible editor: Fumingli)