Ron Kasabian on how to help enterprises build large data Trust

The global economy is still recovering, scandals abound and discord is rife, and in such a context it is difficult to trust one another. So in the past few years, it is not surprising that public confidence in government and business has dropped markedly.

When it comes to the topic of big data, it can create a significant trust problem for the business. The biggest hurdle to big data is not how successful it is, but what makes people really believe in big data and trust big data. And it's not just about believing in the data itself, but believing in the results that can be obtained through data.

Intel Data Center and interconnected Systems division large Data Solutions general manager Ron Kasabian

You know, the value of a single piece of data is limited. The greatest value it brings to the enterprise is to help the enterprise gain insight, grasp the future trend and understand the specific meaning through the associated data points.

However, it is not easy to convince people and trust the insights gained through large data models. The question we face every day is: do we trust our instincts, our experience, our intuition, or our data? Even if the correlation between the data shows sales or efficiency gains, business leaders may not believe what they see in the first time. Therefore, building trust in large data is essential. Unless you can really convince the advantages of big data, people will not give up the way they have been accustomed to making decisions over the years. Therefore, proving the value of large data itself is more difficult than successfully completing a project.

It is not possible to gain trust overnight. This is a lengthy process and must be followed by a strategy in dealing with the relevant business owners. Over the past three years, Intel has implemented large data plans internally and has been very successful. But during this period, we have made a lot of attempts, and have gained a lot of experience and lessons. One of the main problems we face in implementing these plans is to get the trust of relevant principals within the business unit to help them solve the problem.

In this process, we have established six steps to win trust in large data plans.

1. Understand business and data. This may seem necessary, but a complex, in-depth analysis of the business unit requires interaction with key personnel, understanding of their work, collaboration with other parts of the company, and the challenges they face. What are the impediments? What are the factors that prevent them from developing more efficiently? To do this you need a business process planner who must be able to ask the right questions and have an in-depth understanding of the available data.

2. Identify issues and how the data can help. Locate the association between the business problem and the available data. Can this data help solve the problem? At this point you may find that the data you need does not exist. Can you access it? Please note: People tend to view large data as social media and Internet of Things (IoT). They believe that it is necessary to excavate this type of data immediately outside the enterprise, and sometimes this is necessary. But merging external data adds complexity, and I think there is a lot of potential value in the data within the enterprise. Therefore, you need to determine whether the use of external data is necessary, or whether external data can help. Use the internal data first, then expand the scope.

3. Set reasonable expectations--if necessary--to give up. Ensure that the enterprise understand: every solution to a business problem, will be shelved three to four problems can not be resolved. We may spend a few months on some projects, but in the end the results are minimal and even worthless. If the project does not fulfill your expected results, you must learn to give up and then look for the next opportunity.

4. Large data projects have been implemented but are still immersed in old notions. Large data projects are launched at the same time as traditional projects. Business leaders are unwilling to give up familiar processes and technologies, but say "I am willing to believe in data now". You have to prove this to them, but still use their parameters to make decisions.

5. Flexible handling. The Big data analysis you're doing is an exploration. You may find value in unexpected areas. This shows that both methods and tools have great flexibility. Large data tools are still in the early stages of development. You have to be prepared to use large data tools that may be quite different from a year later. You need the flexibility to implement tools, and to upgrade and invest in technology as needed. You may also need to continue to explore the value of large data and present it in a new way to your business, while at the same time you need to master various types of analytical methods and expertise.

6. Focus on results. Sometimes this process can be tedious. You need to focus on the results at this point. Large data is not yet mature and is still in its early stages and there are no sophisticated methods and tools to perform data analysis easily and efficiently. But at Intel, we do everything we can to make use of large data and predictive analysis, which shortens the time for chip design validation by 25% and increases the time-to-market for new chips. These results prove that all these efforts are worthwhile.

Building trust in large data insights takes time. In our case, it took six months just to prove the first business case. But once the plan is in place, companies will start using big data to address major challenges. Not only will the business sector change, but IT departments will also experience changes. As the IT department drives the implementation of large data plans to deliver outstanding results to business units, it will be transformed into a trustworthy partner for the business.