Q&A: Back to the future

Interview with incoming INFORMS President Robin Keller on the past, present and future of INFORMS and its embrace of analytics.

By Peter Horner

INFORMS President-Elect Robin Keller

INFORMS President-Elect Robin Keller

Virtually every dedicated INFORMS volunteer who rises through the ranks to become president of the Institute puts in many years and countless hours of time serving on various committees, editing journals, attending all-day INFORMS Board meetings as an elected officer or, more likely, some combination of all of the above. President-Elect L. Robin Keller, who will assume the presidency on Jan. 1, 2015, is no exception. In fact, one can easily argue that Keller has gone well beyond the call of volunteer duty.

Keller’s membership in The Institute of Management Sciences (TIMS), which merged with the Operations Research Society of America (ORSA) in 1995 to create INFORMS, dates back to 1977. She joined ORSA in 1985. Before and since the merger, she has held more volunteer posts with INFORMS and its predecessors than we can mention here, but a short list would include: TIMS Council member (1991-94), TIMS vice president of finance (1993-94), founding director-at-large of INFORMS (1995-96), past-president, chair-elect and chair of the INFORMS Decision Analysis Society (1998-2004) and editor-in-chief of the INFORMS journal Decision Analysis (2007-2012).

Did we mention that Keller became an INFORMS Fellow in 2004 and received the Kimball Medal for service to INFORMS and the OR/MS profession in 2006?

A professor of operations and decision technologies in the Paul Merage School of Business at the University of California-Irvine, Keller has served as the school’s doctoral program director and associate dean (for the full-time MBA program and for research), as well as the program director for the NSF Decision, Risk and Management Science Program.

Needless to say, Keller brings a lengthy and impressive record of volunteer service to her upcoming stint as president of INFORMS. In her position statement as a president-elect candidate, Keller mentioned several major objectives she would pursue if elected, including: enhance the quality of products and services, manage the scope and diversity of the field, establish a strong and coherent external image of the field, and maintain or improve effectiveness of operations and improve cost efficiency of operations.

In a two-part interview that included a series of emailed questions and answers, as well as a follow-up face-to-face Q&A at the 2014 INFORMS Annual Meeting in San Francisco in November, Keller offered her take on the state of INFORMS, its current activities and the future of the Institute and the profession.

Following are excerpts from both the online and person-to-person interviews.

INFORMS offers many services, hosts great conferences, publishes world-class journals. What sets INFORMS apart from other organizations working in the fields of O.R. and analytics?

We do many of these things well, especially due to our ability to make them valuable and beneficial to our members. Our members make us unique – with their high level of knowledge and capability and willingness to share this with others. We have expertise areas and a breadth and depth within INFORMS that is unique – both across a wide breadth of industry sectors and methodologies.

No organization is perfect. Where would you like to see INFORMS become stronger? What are the challenges INFORMS faces?

Not surprisingly, INFORMS members tend to operate in silos depending on their research or professional interests. Academics obviously know their research area quite well, which is usually some sort of mathematical tool. Industry members are more diverse, but the interaction between those in different industries can be limited. The challenge is: How are we going to connect across the academic and practice divide? What can we do to better serve the people in practice, such as certification and continuing education? To which academics are likely to respond, “I already have certification in my area of expertise, my Ph.D.”

Another challenge is retaining student members once they graduate and move on with their careers, retaining mid-career operations researchers and analysts who move into management positions, and retaining our senior members and retirees. We need to think about cradle-to-grave services. For example, we are working on starting a pro bono O.R. program, a little like “Engineers Without Borders,” where we match people – perhaps senior members or retirees – who want to do pro bono work with organizations that could use their expertise.

The tension, if that’s the right word, between academics and practitioners in INFORMS goes back 60 years to the founding of the two organizations that created it, ORSA and TIMS.

Volunteers, such as many of this year’s Fellows inductees, are crucial to the success of INFORMS and the O.R. profession.

Volunteers, such as many of this year’s Fellows inductees, are crucial to the success of INFORMS and the O.R. profession.

That’s right. Within the INFORMS family, the problem is as old as the hills, and in a way, we’re doing way better now than perhaps we ever did. We created a spring conference that is more appealing to practitioners. Of course, the problem then becomes the academics don’t go to it as much as they used to. We try to tailor, in a marketing sense, what the needs are for each group. However, we have to always make sure that the two groups continue to meet. We don’t want to have a situation where just academics come to the fall meeting and just practitioners go to the spring conference because the two groups have a natural relationship; they need to interact. Academics are producing students and doing research, and they need to hear from industry what kind of curricula industry is interested in and what kind of problems industry needs solved.

Speaking of membership, after many years of flat or slightly declining membership numbers, INFORMS is experiencing a recent spurt in membership growth. Your thoughts?

Yes, we have been growing – this year we are at our highest number of members since 1998. Our student membership is at a record high, and we continue to grow in all our sectors, academic and industry. Much can be attributed to a greater awareness of the value of operations research and advanced analytics and the growing role both play in profitability and decision-making in companies. Recently, U.S. News & World Report listed operations research as No. 2 in their list of top business jobs. As the demand grows, more students are entering our degree programs. And more O.R. and advanced analytics professionals are finding INFORMS to share knowledge and continue their professional growth.

Do you consider membership growth an important measure of success for INFORMS?

Incoming INFORMS President Robin Keller (far right) congratulates outstanding INFORMS student chapter organizers.

Incoming INFORMS President Robin Keller (far right) congratulates outstanding INFORMS student chapter organizers.

I think meeting the needs of the population that we are trying to serve is the best measure of success. Some of those people might not even be members of INFORMS. We are trying to serve the needs of people all over the world. I see it as cradle-to-grave services for people that are interested in O.R. and analytics, with the definition of analytics still evolving. For example, many new analytics programs are coming out in universities. I think our niche in analytics is going to be more modeling-based.

Given INFORMS’ academic roots, is there a ceiling on membership?

In one sense, I think there is a natural limit because there are only so many Ph.D.-level jobs out there, whether it’s in academics or industry. We are reaching out to different fields. We now have journals that publish articles that are far broader than traditional operations research. We offer so many new initiatives such as the spring practice conference, certification and continuing education that appeal to a much wider range of potential members, so there’s more room to grow on the non-academic side.

Practitioners remain a minority in INFORMS. How well is INFORMS serving that particular population?

We designed the spring conference [INFORMS Conference on Business Analytics and Operations Research] to meet the needs of that group and it has been very successful. There are several tracks aimed at practitioners at the INFORMS Annual Meeting in the fall. INFORMS offers a whole suite of services that should be attractive to those in the practice field such as the CAP® [Certified Analytics Professional] program. Nearly 200 people, many of them practitioners, have already earned certification and many, many more are expected to become certified in the future. Practitioners can use INFORMS’ Analytics Maturity Model to rate how well their organization is doing compared to others and perhaps use the findings to recommend to their boss that they need more analytics-oriented people in certain departments. Practitioners have benefited from our continuing education program.

INFORMS is built on a three-legged foundation of key activities: journals, meetings and subdivisions. How are these areas doing?

We are breaking participation records in our journals, meetings and subdivisions. Last year and this year our journals received the highest levels of submissions ever, and we have our highest number of members participating in our subdivisions. And this annual meeting is the largest in our history. We are launching our 14th scholarly journal, Strategy Science, following upon the recent addition of Service Science. Having served as the editor-in-chief of Decision Analysis, I know first-hand about the great support provided by the staff and volunteers of INFORMS in producing our journals and the very valuable impact on careers that our journals provide.

Certainly this meeting is a huge success in terms of attendance with nearly 6,000 attendees, an all-time record. Is the annual meeting reaching a tipping point? Can it become too big?

I don’t see it as a problem. Again, the fall meeting is an academic-style conference, and the vast majority of attendees are giving talks. The ratio of attendees to presentations is roughly one to one. At a meeting like this, it’s really important to find your interest group and the sessions that meet your academic, research or professional interests. Many of our subdivisions are organizing their own focused, stand-alone meetings. It’s become very popular to attend those smaller conferences as well as the big annual meeting.

How is INFORMS doing financially?

We have worked hard to ensure a strong and healthy INFORMS. From 2011 through 2013, we ended each year with a surplus, and we expect another surplus this year. We have been focusing on building the resources needed for us to invest in the future of INFORMS. For example, we’re investing in programs that increase member benefits and value such as INFORMS Connect, high-quality continuing education, certification, new meetings and new networking opportunities. We also want to focus on increasing awareness of the profession and providing stronger links between academics and industry.

INFORMS has launched several huge initiatives in recent years, including the CAP program, continuing education, the Analytics Maturity Model, etc., all of them tied to analytics. How are they contributing to the success of INFORMS?

These analytics initiatives allow individuals and organizations to gain added analytics and O.R. knowledge and to demonstrate their proficiency. For those not currently in a degree program, the CAP certification provides a route to enhance skills and document the skill level. This leads to increased visibility of O.R. and analytics, and an increased awareness of the value and capabilities that advanced analytics and O.R. provide.

I understand the numbers for CAP and continuing education are relatively modest at this point.

I think certification appeals more to non-academic professionals, people who don’t have Ph.D.s or professionals who may be interested in a job change. Independent consultants and younger people within an organization who are trying to move up the corporate ladder are good candidates for CAP. Those people aren’t necessarily INFORMS members, so it’s a little harder for us to reach them. One way we could do that is through these new analytics programs that are popping up in business schools. Perhaps graduating students could launch their business or consulting careers with CAP certification. It wasn’t originally designed that way since CAP includes an experience component, but that’s something we could look into.

Likewise, with continuing education, we’ve designed some excellent courses that have proven quite popular with the people that have taken them. The course designers would love to get the word out there even more and have more people take the classes.

You can’t expect overnight success with new initiatives such as these. It takes time to develop demand, but I’m confident it will happen. It’s the same way when we start a journal. You just have to nurture your baby as it’s growing.

“Operations research” and certainly “analytics” – however you define it – are gaining more recognition in the corporate sector. The U.S. News & World Report story mentioned earlier is just one example. Where do you see O.R. expanding in the corporate sector?

Our techniques are already widely used across industries and functional areas, from decision analysis in pharmaceutical and oil and gas industries, to mathematical programming for logistics and transportation, and on to game theory to gain insights for operational strategies, such as random scheduling for the security patrols at Los Angeles International airport.

The 2014 INFORMS Annual Meeting drew record crowds, including this overflow crowd for a panel session.

The 2014 INFORMS Annual Meeting drew record crowds, including this overflow crowd for a panel session.

As just one example of a novel use of operations research that has come of age in the past couple of years due to advancements in information technology, Professor John Turner, my U.C. Irvine colleague, developed a mathematical programming model to help with dynamically placing display ads into online video games. Recently, John has been looking at friend links in social networks and developing mathematical programming models for strategically placing ads in such venues.

We will continue to use our O.R. foundation of understanding a business problem, modeling it and drawing prescriptive insights.

Do you see analytics, especially advanced analytics, increasing awareness of O.R. and its value in business and operations?

The member survey done in 2011 and again in 2013 asked this question and a strong majority agreed. It will be important for us to emphasize that we do advanced analytics, and, in particular, we promote prescriptive analytics to aid decision-makers to find better ways to do what they do now or what they might do in the future with advanced IT capabilities.

What’s the one thing you would want the INFORMS membership to know or understand about the Institute that many of them probably don’t?

INFORMS is like an elephant being approached from different sides by different people. We are large and have very many different faces. Most people who are tightly engaged with INFORMS see it primarily from one perspective. Academics may first notice the scholarly journals, then the fall conference. Students see the conferences and the career services. Practitioners see the spring conference, CAP certification, OR/MS Today and Interfaces and continuing education. INFORMS Board members emerge from their base within INFORMS, then get a generalist’s perspective and focus on strategic plans, budgets and working with our outstanding staff.

So it is important for everyone to try to understand the different objectives of the different stakeholder groups, so we can work together to meet our multiple objectives.

You’ve been an active member of INFORMS and its predecessors for many years and served in many leadership roles. How did those roles prepare you for the presidency? What motivates you to spend so much time and energy on the Institute? How important are volunteers to the Institute?

Leading the Decision Analysis Society of INFORMS gave me the perspective on organizing conference clusters and building volunteers’ skills to grow into new volunteer positions. Working on the campaign to found the Decision Analysis journal and later serving as its editor-in-chief gave me perspective on publications. My grandmother advised me to change volunteer jobs every three years, and I try to follow her advice.  Three years is time enough to learn the role, come up with some innovations and train a successor. I try to volunteer for jobs that I like, both as a professor and within INFORMS.  I also stick to my plan to rotate out of a job before my energy wanes.

This is a volunteer organization, and our volunteers are the lifeblood of all that we do, working alongside our skilled professional staff.

Fun question: Assuming you have time away from your extensive academic, research, professional and volunteer responsibilities, what do you do for fun?

I like to swim. I enjoy reading books. My husband and I like to visit his family farm in Iowa and my sister’s horse ranch in Colorado. I like the outdoors, so we go for walks on the beach near our home in Corona del Mar.

What are your priorities during your year as president? Do any stand out as more important to you for INFORMS?

It is a privilege to be in this role representing all our members and the field as a whole. I want to help people in their careers and help people in organizations get what the organization needs.

A year from now, how will you measure your success as president of INFORMS?

Many times, when we say later that someone did a good job, they might have done a good job with a super-confidential serious issue. The strange thing is that you could do a really good job that you wouldn’t even publicize because it solved a problem with strongly held opinions on different sides and that you kept things running almost without the regular members knowing that there was even a problem. Hopefully, we won’t have any serious problems like that, but if we did, I would think I would be pretty good at trying to understand the different sides and preventing something from escalating into a bigger problem.

Finally, how important is the analytics movement to the future of INFORMS?

I would say the analytics movement is really going back to the roots of O.R. That is, find real problems, understand the real problems, be sure you have the soft skills to explain them, and make sure you have the O.R. and analytics skills to analyze what to do and implement it.

Maybe as individuals and as an organization we forgot somewhere along the line to emphasize that whole dimension, and now by emphasizing analytics we’re getting back to it … back to the future.

Peter Horner (horner@lionhrtpub.com) is the editor of OR/MS Today.