Share with your friends


Analytics Magazine

Analyze This!: Virtues of agile development methodology

Vijay MehrotraBy Vijay Mehrotra

A couple of years ago, I read an article by Erick Wikum entitled “Making analytics work through practical project management” [1] that stuck with me for many months. I recently spent some time discussing the relationship between project management and analytics with Erick, the new president of the Analytics Society of INFORMS (congratulations, Dr. Wikum!).

At the outset of our conversation, we both acknowledged a fundamental appreciation for some of the basic concepts that have long been codified in the Project Management Body of Knowledge (PMBOK) [2]. Erick explained that this respect was a direct result of his experience working in various internal operations roles after finishing graduate school. Often times, he explained to me, projects simply emerged organically from casual conversations that concluded with someone promising to “have someone work on that.” In many such situations, he observed that there were no timelines and no clear deliverables, but rather just smart people “doing stuff.”

As a result, technical professionals would frequently find themselves working on “projects” for which no one could accurately answer basic questions like, “What are you trying to accomplish?” and “What does success look like?” Only after working for several years as an external consultant, including stints with IBM Global Services and Tata Consultancy Services, did he come to truly appreciate the value of things like project proposals, statements of work and project plans.

At the same time, Wikum also pointed out some of the limitations of the PMBOK methodologies. “There’s a certain amount of invention in almost every analytics project, although we are never sure up front just how much,” he noted, “and so estimation is more art than science.” From my perspective, this is because our work almost always involves touching a lot of other systems and people, including data sources, subject matter experts, business users and work systems [3] that utilize our output. In a subsequent email exchange, Erick wryly noted that “until you marry data to math, you don’t really know what you’re going to get.”

Despite all of this, most individual managers and organizational cultures are quite averse to uncertainty and ambiguity, and so we are often asked to play along when planning projects. But at the outset of a project, our knowledge of what is truly required is often limited and as such the risk inherent in our estimates and plans is at its highest. Sadly, when following the traditional “waterfall” project management methodology, this is the period during which project requirements and specifications are finalized and the project budgets are established. As a result, individual and organizational mindsets are often prematurely anchored to highly imperfect estimates of what is to be done, how long it will take, how much it will cost, and what it will deliver.

Though Wikum and I were discussing predictive and prescriptive analytics projects, many of the same project management challenges are also present in complex software development efforts. As a result, long before the general public had become fascinated with analytics and data, people in the world of software development were wrestling with many similar issues. In 2001, a small group of software development thought leaders released the now-famous Agile Manifesto [4], which has in turn propagated into the “agile development methodology,” an increasingly popular approach to managing projects inside and outside the world of software.

Much has been written about the agile development methodology [5] over the past two decades, far more than I have either the knowledge or space to fully describe. But here are some of the highlights from my perspective. Agile development processes are organized around an iterative series of rapid cycles (often referred to as “sprints”) that are driven by high-level goals and clear priorities while producing a working solution at the end of each sprint. Key attributes of agile include: (a) the delivery of incremental value at the end of each cycle; (b) intense communication and collaboration between developers and stakeholders throughout, resulting in a constantly improving (and shared) understanding that serves to guide future decisions; and (c) the ability to easily accommodate changes to goals and requirements in future sprints as a result of what is discovered along the way.

During our conversation, Wikum provided vivid illustrations of some of agile’s key virtues. One of the projects he described to me featured a simulation-optimization model. While there had been a significant discussion of the project priorities at the outset, early iterations produced model prototypes that gave the project’s business customers a much better understanding of what was possible, which in turn led to a shift in priorities and design choices.

“People don’t always understand or respond to the same words in the same way,” Erick observed. “Sometimes clarity only emerges when there is something to see, touch and feel.”

On another project that he mentioned, an early sprint that required the team to build a working model and evaluate its output proved to be very valuable. In particular, this initial build helped the data science team to arrive at a far deeper understanding of the input data than they had had after first examining the data dictionary. “Some of these kinds of details are just too complex to explain as clearly as you would like,” Wikum pointed out.

Finally, Erick pointed out the importance of continually prioritizing the most important elements of a project and managing future decisions with those priorities clearly in mind. “Depending on budgets, in the end you may not get all of what you want. But if you are taking an iterative approach, you should come out with something that works and has value.”

Given that more than half of analytics and big data projects ultimately fail to achieve their goals [6], this last point is perhaps the strongest argument of all for the agile methodology and mindset.

So given all of this, why doesn’t every organization use agile to manage every analytics project? Well, this sprint is over, but that’s a great question for a future iteration of this column.

Vijay Mehrotra ( is a professor in the Department of Business Analytics and Information Systems at the University of San Francisco’s School of Management and a longtime member of INFORMS.


  5. A good place to start is here:

Related Posts

  • 70
    The CUNY School of Professional Studies is offering a new online master of science degree in data analytics. The program prepares its graduates for high-demand and fast-growing careers as data analysts, data specialists, business intelligence analysts, information analysts and data engineers in such fields as business, operations, marketing, social media,…
    Tags: data, analytics
  • 65
    Benjamin Franklin offered this sage advice in the 18th century, but he left one key question unanswered: How? How do you successfully drive a business? More specifically, how do you develop the business strategy drivers that incite a business to grow and thrive? The 21st-century solution has proven to be…
    Tags: data, analytics
  • 65
    On one end of the spectrum, labeled “Do Stuff,” organizations focus on action taking, a laissez-faire approach to project management, with little documentation and loosely defined deliverables, timelines and budgets. On the other end, labeled “Buttoned Up,” organizations take a disciplined approach to planning, monitoring and executing projects, with documentation…
    Tags: project, analytics, management, data, projects
  • 61
    September/October 2018 Analytics revolutionizing healthcare Values that should guide AI revolution The future role of AI in fact checking Data science at Monsanto Strengthening global supply chains July/August 2018 Why optimization models fail AI: Path to an intelligent enterprise How to truly listen to customers Close the deal: marketing &…
    Tags: analytics, data
  • 57
    Most business leaders today believe in the value of using data and analytics (D&A) throughout their organizations, but say they lack confidence in their ability to measure the effectiveness and impact of D&A, and mistrust the analytics used to help drive decision-making, according to a new survey from KPMG International.
    Tags: analytics, data


Using machine learning and optimization to improve refugee integration

Andrew C. Trapp, a professor at the Foisie Business School at Worcester Polytechnic Institute (WPI), received a $320,000 National Science Foundation (NSF) grant to develop a computational tool to help humanitarian aid organizations significantly improve refugees’ chances of successfully resettling and integrating into a new country. Built upon ongoing work with an international team of computer scientists and economists, the tool integrates machine learning and optimization algorithms, along with complex computation of data, to match refugees to communities where they will find appropriate resources, including employment opportunities. Read more →

Gartner releases Healthcare Supply Chain Top 25 rankings

Gartner, Inc. has released its 10th annual Healthcare Supply Chain Top 25 ranking. The rankings recognize organizations across the healthcare value chain that demonstrate leadership in improving human life at sustainable costs. “Healthcare supply chains today face a multitude of challenges: increasing cost pressures and patient expectations, as well as the need to keep up with rapid technology advancement, to name just a few,” says Stephen Meyer, senior director at Gartner. Read more →

Meet CIMON, the first AI-powered astronaut assistant

CIMON, the world’s first artificial intelligence-enabled astronaut assistant, made its debut aboard the International Space Station. The ISS’s newest crew member, developed and built in Germany, was called into action on Nov. 15 with the command, “Wake up, CIMON!,” by German ESA astronaut Alexander Gerst, who has been living and working on the ISS since June 8. Read more →



INFORMS Computing Society Conference
Jan. 6-8, 2019; Knoxville, Tenn.

INFORMS Conference on Business Analytics & Operations Research
April 14-16, 2019; Austin, Texas

INFORMS International Conference
June 9-12, 2019; Cancun, Mexico

INFORMS Marketing Science Conference
June 20-22; Rome, Italy

INFORMS Applied Probability Conference
July 2-4, 2019; Brisbane, Australia

INFORMS Healthcare Conference
July 27-29, 2019; Boston, Mass.

2019 INFORMS Annual Meeting
Oct. 20-23, 2019; Seattle, Wash.

Winter Simulation Conference
Dec. 8-11, 2019: National Harbor, Md.


Advancing the Analytics-Driven Organization
Jan. 28–31, 2019, 1 p.m.– 5 p.m. (live online)


CAP® Exam computer-based testing sites are available in 700 locations worldwide. Take the exam close to home and on your schedule:

For more information, go to