After "Madly in Love", examine the pain and pass of big data

Like a high bonus car Grand Prix, first prepare the car, familiar with the track, and then adjust the good mentality, timely sprint, to get a good position. The big data also has the gestation period, the preparation period, the result originates from the demand and the desire. Each period has pain points and difficulties, the result is the best.

Big data has big energy. It is significant for economic growth-"in the US, in retailing and manufacturing alone, big data analysis has the potential to raise GDP by 325 billion dollars." "This is an exciting estimate from analysts. At a time when global economic growth is slowing and even the state of weakness is desperately in need of new growth, people are looking to the big data that is expected to take this responsibility.

Pragmatic and savvy entrepreneurs remain sane and sober while being surrounded by enthusiasm and passion for big data. They need far more than a flowery description of the big data outlook, and they need to be persuaded not to turn into a data-driven enterprise as quickly as possible, and to show how big data can be applied to their businesses, demonstrating the benefits of big data.

The inescapable reality is that big data also has pain points, and these pains have become troubling stumbling blocks. The so-called "general rules do not hurt", these pains are "doctor" gave a good prescription. Large data efficiency is the most direct and powerful argument. By dissecting every implementation experience, you will find that the answers that entrepreneurs need are included. To reason, to emotion, they can no longer resist the temptation of large data.

Large investment data have guidelines

In an interview with Fortune, Fille Macavetti, chief brand officer of Starwood International Hotel Group, described the big data as: "There is a passageway between the general manager's office and the front desk in our oldest hotel." Through this channel, the general manager can see every new guest and greet them like an old friend. I think the big data is the 21st century channel. It gives us a better understanding of the guests and provides the customer with the services he really needs. ”

Matt Asay,nosql, one of the best-known companies in the field of database technology, MongoDB, a vice-president of the company, once exclaimed: "The attraction of large data is like gravitation, so strong that people can't escape, people can't help but want to carry out related projects." ”

The CIO of more than one large enterprise publicly stated that large data saved the company a lot of money--1% 's efficiency has brought in millions of of billions of dollars in spending cuts, and that the effective mapping of sales to customer databases has resulted in 10% sales growth ... These mouth-watering data are consistent with Bain&company's findings: Companies that use big data analysis are far ahead of rivals – the likelihood of a 25% increase in their financial position in the industry, four times times faster, and a three times-fold execution effect; The likelihood of using data support decisions is twice times higher.

As a then, companies have done something. A Gartner survey showed that 64% of surveyed companies said they were deploying or planning to deploy large data projects. It is embarrassing, however, that 56% of respondents are confused about how to get value from big data, and 23% of respondents have doubts about the exact definition of big data.

"The big data boom doesn't mean that any business needs to jump into the big data wave," he said. "In the view of Jamal Khawaja, director of Infinitous Global Services, the adoption of large data is a long-term and multidisciplinary work that is not a small test of the enterprise's platform, service and internal investment capabilities."

Based on years of experience in enterprise research, he has given the four criteria for determining whether an enterprise is in the best time to launch large data: There has been a certain accumulation in business intelligence (BI), the amount of data in hand has reached a certain scale, corporate culture is recognized, with sufficient talent reserves.

"Only by properly processing the existing data can we find a way to solve the problem from the discrete data stream." There are two criteria to measure whether the data is handled properly: first, it brings the professional skills required for a higher level of analysis, and the second is to obtain some financial benefits. "Jamal Khawaja in detail," the data in the historical dataset can offer many opportunities, such as excluding running exceptions. Before making big data investments, it's important to be aware that big data investments are different from other investments, that they are not driven by existing, obvious problems, that you don't have the standard ROI or NPV model for big data investments, and that the expected return is not obvious. It can be said that investment in large data is inherently opportunistic. Therefore, there must be a corresponding corporate culture to accept this. The importance of professionals such as data scientists is needless to say, and the way they extract value from the data is dazzling. ”

If the examination is complete, feel confident, can practice the big Data project, still cannot be taken lightly. Myles Suer, senior manager from Informatica, points out that many of the big data projects are actually just "old" versions of previously existing BI projects, and that the applicants do so only to ensure that the funds are in place or the project is approved, which means that the project argument must be rigorous.

"In addition, we need to be wary of past mistakes in it." Myles Suer says it has undergone three major stages of development: local systems, ERP, bi/large data, and "ERP should have addressed all local solutions, but it only provided transaction information." We want to avoid bi/big data. Otherwise, CIOs are likely to hear CEO/CFO's complaints that the accumulated data does not make the company more money. Connecting trading systems, BI, and planning systems with large data is a viable way to bring more profit to the company. ”

The first will climb, farewell to go, to run. On the big data can also not be impatient, also can not be anxious-the implementation of large data like to participate in a high bonus car Grand Prix, to prepare a good car, familiar with the track, adjust the mentality, timely sprint, finally to get a good position.

Risks that cannot be ignored

"Some companies may underestimate or completely ignore the risk of some big data if they are in a hurry to adopt the methods and techniques of large data." The implementation of large data is equivalent to a big gamble for companies, and no matter when the business must not be taken lightly. Understanding these risks and making plans to deal with them is critical to companies that are likely to benefit from large data. Rick Delgado, a well-known IT management consultant, sounded the alarm.

In his view, the biggest risk in implementing large data is security. The more data the system processes, the greater the likelihood that the data will be lost. "Recent data leaks by target companies and Coca-Cola companies show that data loss incidents are imminent and the costs are enormous," he said. "Rick Delgado stressed that the importance of IT security is what many businesses lack, which can lead to much scarier results than ever before in the big data age."

Privacy concerns are also a hurdle to go around. It is both a legal issue and a commercial issue. There is a nagging debate about data ownership, especially when it comes to consumers and businesses, or businesses and big data providers. Companies using the data they collect about users can lead to some legal consequences. Misuse of the data may lead to lawsuits, fines, and even further serious consequences such as industry regulation. A typical example is that the NHS hospital was accused of going to court for selling patient information to an insurance company without permission. In addition, while some breaches of privacy may be technically legal, companies will still be at risk of a major blow to their reputations.

Using large data another risk that cannot be overlooked is that it can cause a loss of agility to an enterprise. This may sound strange, because big data is meant to help companies respond more quickly to real-time data, but don't forget that different datasets of large data are always placed in different platforms that use different software programs. Large data requires that data be managed, organized, and stored in an extremely efficient manner so that appropriate analysis can be carried out and action taken. If everything is disorganized and there is no reasonable management of input data, it is easy for businesses to be paralyzed when they need to act quickly. No surprise, then, that the company loses its preemptive chance in the "Time is money" market competition.

In addition to poor data management, the risk of distorting data is noteworthy. With big data, companies sometimes take it for granted that it can do almost anything, but it's not that simple. Big data can show what's going on, but it doesn't give an explanation of why things happen. To get these explanations, it is necessary to make a correct analysis, but sometimes it is difficult to rely on the data itself. For example, the data show that at some time of the year the sales volume blowout, but why blowout must be interpreted by specialized data experts, or the conclusion may be wrong, resulting in ineffective action. In the long run, the distorted data will end up wasting the company's wealth in vain.

How do companies reduce the risks associated with big data? Rick Delgado the answer: hire professionals who are good at managing large data, analytical and capable of getting the right conclusions. Having the right people in key positions will help businesses make efficient use of data and reduce the chance of errors when they need to interpret data and use the latest large data tools. Companies should also routinely conduct self-assessment to ensure compliance with current privacy regulations and commercial regulations. It is also important that companies be honest and transparent, so that customers know how the enterprise uses their data, which will help customers trust and loyalty.

In the face of potential risks, Rick Delgado is still cautious and optimistic: "There is no doubt that big data can provide huge benefits." But companies should take a pragmatic approach to risk. No new technology or practice is risk-free, and companies can be prepared to respond to challenges as long as they keep this in mind. ”

The harmony between the CXO

By 2015, Gartner says, 1/4 of large companies will have the post of Chief data Officer (CDO). According to Cloudpro.com, there are now 100 CDOs in place, 65% in the United States and 20% in the UK.

Gartner describes the CDO's mandate: "They are responsible for data-related core processes, data management, and coordination of data use." This is similar to the CFO, which controls the financial process and makes the money flow reasonably within the enterprise. ”

For example, the Bank of America, which took the initiative to move ahead of the trend, began its CDO, John Bogetta, in December, before the big data became the focus of CIO openings. His main responsibilities are data management and quality control.

As CDOs have sprung up, have they grabbed the exclusive territory of the IT and the CIO? What CxO, such as the Chief Digital Officer (CDO), CTO and chief Marketing Officer (CMO), need to adapt to the big data age?

After a lot of interviews, Harvard Business Week found an embarrassing fact: CxO not around the core of large data, especially CMO and CIOs are common problems in businesses seeking to benefit from big data.

Indeed, there are differences in the ways of thinking of various cxo and the pursuit of performance goals. But don't forget, research has shown that data-driven enterprises are "5% more productive, 6% more profitable" than non-data-driven enterprises. Productivity and profitability are the highest goals for a company. As a result, CXO must be aware of the fact that big data is the perfect weapon to help policymakers sift, analyze, and predict when markets need to innovate.

"Today, the role of technology in the business process has changed dramatically, and these changes require CXO to collaborate." "said John Dodge, a well-known it media man.

Matt Ariker, chief operating officer of the McKinsey Consumer Market Analysis Center, gave the pertinent advice to improve CXO's fragmentation: companies must have an effective decision-making framework, and CXO's leadership team must support each other. This affects every step of translating data into value, including developing strategies, choosing usage scenarios, budgeting, and deployment. Clear requirements are particularly critical, requiring CMO to identify business objectives, usage scenarios, and specific requirements related to data and analysis, CIOs need to provide feasibility and cost analysis based on usage scenario requirements, CTO provides strong technical support, and CFO prepares "wallet". These need to be weighed and chosen in terms of cost, time, and priority. Next, the formation of a suitable and diverse team, the process of transparency, regular follow-up, which is essential for effective communication, if necessary, even need to "translate", so that CxO can "understand" each other's words. For example, CMO needs people who know both the customer and the business and have geek thinking, and the CIO needs to communicate with the technicians who have a deep understanding of the marketing activities and business. Then, the leadership team should be step-by-step, can first carry out a number of reference work to test team collaboration and new processes, and then spread.

The effective use of large data has become a watershed in many industries to distinguish winners from losers, but there are no shortcuts for companies to go. No single CXO can enjoy the success of large data alone or the failure to bear large data alone.