Big Data: Why Your Project Needs It

Beyond the technological requirements, transitioning organizations face two challenges: necessary level of effort and standardization. What are the lessons learned and where can PMOs find the resources they need to mine the value of big data?

Big Data is here. Concise, targeted decision-support information can be extracted from colossal electronic warehouses of data. The challenge to businesses is to develop methods that mine true value from terabytes of data. Among large and small organizations, U.S. government agencies such as the National Science Foundation, Department of Defense, and U.S. Geological Survey have all placed heavy bets on the success of Big Data projects.

Big Data is frequently perceived as an information technology problem. However, this perspective can cause some costly missteps. While compiling massive data stores is often treated as an IT project—and IT staff will be instrumental in assembling the mechanisms—for a Big Data implementation to meet organizational intent and thrive, it must be sponsored and driven strategically by a coalition of management functions. IT is not the ultimate consumer of the information. Having all stakeholders invested early in the data storage and retrieval design process will ensure that the system will ultimately serve users with the best and most efficient level of decision support.

For project managers, Big Data can represent a gold-mine opportunity for performance analysis and benchmarking. Although all projects possess unique elements, organizational tendencies are often cyclical and proper data mining and trend analysis can uncover correlated elements that are not apparent in traditional, smaller-scale data analysis.

Already, project-oriented Big Data research has identified a compelling relationship between the integrity of the project baseline plan and the eventual success, or failure, of the project in execution. By measuring a small number of quality indicators at project initiation and comparing them to successful and unsuccessful past projects, the system can provide solid foresight into the eventual success of a new project.

A natural reaction for a program manager may be to shy away from predictions derived from the work of others; however, the advantages are too great to ignore. For example, a program manager might use this intelligence to gain the resources needed to ensure project success by demonstrating how these additional resources have increased the likelihood of success in the past.

In order to maximize the benefits to the program management office, program managers should:

Actively engage in the design and implementation stages of the Big Data project. Stakeholders must define and communicate specific goals for data analytics.

But, keep it simple. Start with small goals, such as ‘understanding forecast trends.’ Add loftier goals as intermediate achievements are made.

And finally, actually implement changes based on the information provided from the system. Optimize plans to reduce the likelihood of repeating failures.

Stakeholder input aside, the final design must have flexibility in mind. Spending too much time on the initial staging of the hardware and information infrastructure can shortchange long-term flexibility. The system must have the ability to evolve in phases, increasing in usefulness as the system ages while perpetually collecting up-to-date information.

Big Data is the cornerstone of true business intelligence. The intelligent system can summarize organizational tendencies rapidly and on-demand. New business proposals can be created in a swift, realistic fashion by utilizing accurate historical performance. Proposals become highly defensible, built upon a foundation of aggregated real-world events. While success is never certain, reliance on objective information gleaned from past successes and failures will drive accurate planning.

Wrangling historical data is not always easy or inexpensive. However, evolutionary normalization of data and capability upgrades, performed in several phases, allows early benefit from Big Data while configuration continues for deeper analysis. Early on, by simply identifying and cataloging baseline budget, intermediate forecasts, and final actual cost for all projects, a new early-phase project can immediately be analyzed and compared to historical peers at their early phases of execution. In one instance, management may view the cost performance as weak, but the program manager can demonstrate that cost efficiency exceeds the performance of 75% of historical projects. With an even more advanced system, management could tie together the integrity of the project baseline, early cost performance, program risk analysis, and historically similar programs to develop a reliable optimization strategy for the project from the outset, ensuring on-time delivery while minimizing resource usage and maximizing profitability. Big Data is here to stay and PMOs would do well to take full advantage.


Thomas Polen is director of professional services for Acumen