Any reseller that has been in business knows technology platforms come and go, but the value behind the technology persists. Rather than making the error of chasing the latest hot technology, smart resellers sell the value proposition behind the technology. With big data, for example, it’s important to understand how to use the tools required to complete a big data project, but the insight gained from the project itself is where the real value lies. Understanding which big data skills will deliver ongoing returns is more important than mastering the latest Hadoop techniques.
As we have noted in the past, there is an ongoing shortage of data scientists to support big data initiatives, so skilled data scientists can still write their own ticket. However, the big data skills gap is shifting as programming trends are changing. According to a survey by CrowdFlower Inc., demand for NoSQL database skills is declining, while demand for RDBMS is on the rise. Because SQL is well-established, there is growing demand for SQL on Hadoop for big data. CrowdFlower's Justin Tenuto adds, "Hadoop, Python, Java, and R round out our top five in-demand skills."
While demand for expertise with different programming and database languages comes and goes, strategy behind the implementation tools has lasting value. When you consider the value of big data in the context of making better business decisions, not just creating better big data projects, you gain an appreciation for the big data skill sets you truly need to be successful.
Here are just four of the big data skills that are sure to stand you in good stead, no matter what technology you use:
1. Developing big data use cases
Don’t think of big data as a technology in search of a problem. Rather, start with the real business problems that your customers face and then determine if big data can provide insight. Take the time to understand your customers’ goals and issues and any concerns that may impede a big data initiative, such as security risk. You may have to get the IT team to agree to adapt its security to take advantage of data that will have value for the C suite.
Once you understand the goals and the impediments, you can start thinking about developing use cases. Use cases require both business acumen and technical prowess. A successful use case starts with company objectives and then assembles the data necessary to deliver insight about those objectives. The most typical big data use cases tackle challenges such as optimizing business processes, addressing security issues, understanding customer behavior and assessing new products and markets.
Being able to work with clients to help them articulate their business objectives and then translate those objectives into a big data initiative is a skill that will never go out of style.
2. Applying data warehousing best practices
Expertise with MySQL, SQL Server, DB2 and other relational database languages is certainly valuable, as is an understanding of NoSQL, massive parallel processing, data warehousing, etc. However, understanding how to apply Hadoop for big data storage and processing is not a substitute for proven data warehousing and storage management techniques.
The real skill is knowing how to apply the right data management platform for the big data task at hand. To make the most of big data requires data stewardship. Data quality and integrity have to be maintained, and interoperability between applications has to be carefully managed.
3. Delivering data-driven insights
The ROI from big data comes from turning insight into action. The skill here is knowing how to identify and gather together available data sources, including finding those silos of valuable data hidden in the organization. Assembling the right data and then applying the right questions to uncover actionable insights is part science and part black art.
This is where expertise in both database technology and best business practices becomes invaluable. If you can combine the right use case with the right data sources and apply the right analytics, you can deliver results that are truly valuable.
As part of delivering data-driven results, you also have to demystify the data. That requires more than extracting the information; you also have to find ways to make it easy to understand. Using MapR and other tools will help, but you have to be able to build the right user interfaces and reports to present the end results so that the results point to a clear plan of action.
4. Mastering soft skills
In many ways, big data analytics is more about people than technology. Big data initiatives are a big investment for most organizations, and at the outset, results are largely unguaranteed. Company executives are taking a risk as to whether or not big data will deliver answers to questions that have yet to be defined. There is a leap of faith required in most big data projects, so you need to inspire trust and help customers make that leap.
Big data success requires good listening and communication skills. More than other types of data projects, big data is collaborative. This means understanding how to work as a team and to guide and nurture the process toward a positive outcome. You need to be able to work with others, understand their needs and concerns, and appreciate how to assuage those concerns and address those needs in a calm and reassuring manner.
When tackling any big data initiative, or any technology project for that matter, you have to look beyond the process and start with the desired end result. Use your business savvy to develop a strategic road map to clearly define objectives and outcomes and then apply your technical expertise to deliver the results. The consultative value of knowing how to apply technology to business has more value than technology alone.