Without a big data roadmap, your customers are likely to get lost. The world of big data is still a big unknown for many IT professionals and CIOs, and without the proper guidance your customers could spend a lot of unnecessary time and money.
Big data projects often fail because of inadequate preparation. A CIO survey by InfoChimps reveals that while 81 percent of CIOs listed big data as one of their top priorities, 58 percent say big data projects fail because the scope of the project is inaccurate. Other reasons cited for big data failures include lack of business context (51 percent) and lack of expertise in interpreting the data (51 percent). All these problems can be eliminated with a well-defined big data roadmap:
Step One: Define the problem. It’s always best to start at the beginning. Some organizations like to embrace the next cool technology without fully understanding the implications. Make sure senior management understands that big data adoption will affect the company’s overall data management strategy. Determine what the objective of the project truly is before you start, and ascertain what value they expect to get from big data.
Step Two: Create an interdisciplinary team. Since big data is a company-wide initiative you need to get buy-in and participation from stakeholders across the organization. Data tends to reside in department siloes so you want to make sure you have access to all relevant data. And you want to make sure that you get the right perspective from each department to extract the right information. If you are looking for marketing insight, include marketing; if you are looking ways to improve products or services, including product management.
Step Three: Inventory your resources and determine costs. Once you have determined your big data objectives and established a team, assess your data and IT resources. Do you have access to the right data from both inside and outside the company? Do you have adequate data storage? Remember that you could be handling petabytes of data and your storage needs will only grow. Also inventory your available skillset to make sure you have the right resources to manage the architecture, the programming, and analytics. For example, do you need to hire Hadoop experts? Once you have an idea of the resources available and what’s missing, you can create a more accurate budget.
Step Four: Select a proof of concept. Before you launch a larger big data initiative, start with a use case. This is where your interdisciplinary team is invaluable. Brainstorm and pool different perspectives to determine what’s possible and what resources will be needed. The use case is an important step in your big data roadmap since it gives you a chance to create, test, and refine the end-to-end process, from the initial data gathering to analyzing the results. After the first use case, you will have a better understanding of the potential pitfalls and hidden costs, as well as the potential rewards of your big data initiative.
Step Five: Share and results. Since big data is going to have an impact on the entire organization, share the results as soon as possible. This will promote a “big data culture” and give other managers in the organization a chance to see the impact and value of big data. The goal is to create an “information-centric” culture where everyone understands the potential value and applications for information, and big data receives senior management buy-in.
Step Six: Assess the ROI. Now that you have the results from your use case and feedback from the organization take a hard look at the return on investment. Has your big data roadmap led you astray? Will you be able to apply your use case findings in a specific way that generates profits or saves costs? Do the findings promise to pay for the project over time? The results from your pilot project should provide some clear indicators as to whether expanding your big data program will yield positive results.
Step Seven: Rinse and repeat. Now that you have gotten your feet wet with a pilot project, you are ready to expand and improve the big data process. Assess your team’s strengths and weaknesses and expand as needed. Add new data sources. Upgrade your infrastructure to accommodate more data and more data processing. Make sure you have the right kind of data storage. Refine the big data process so it can be adapted to new questions to deliver new insights.
If you work with customers to define their objectives in advance and take a hard look at the results at each step of the way, then the big data roadmap will lead to new profits for both you and your customers. Where do you think you bring the most value in the big data process?