Q&A: The state of INFORMS and the profession
In a wide-ranging interview, Steve Robinson offers his view on analytics, new initiatives, diversity and more as he prepares to take the reins as president of the Institute.
By Peter Horner
Stephen M. Robinson, professor emeritus of Industrial and Systems Engineering and of Computer Sciences at the University of Wisconsin-Madison, has served INFORMS and its predecessors, ORSA and TIMS, in various volunteer roles throughout his long involvement with the Institute. Those roles included head of publications for ORSA and TIMS, and secretary, treasurer, president-elect and, as of Jan. 1, 2014, president of INFORMS. I sat down with Robinson on the final day of the INFORMS Annual Meeting in October to get his thoughts on a wide array of topics, starting with INFORMS initiatives new (CAP and continuing education) and old (journals and meetings). We also talked about his goals for the upcoming year, volunteerism, diversity, finances and, of course, INFORMS catching the analytics wave.
Following are excerpts from the interview.
What do you absolutely want to achieve during your year as president?
The one thing I want to do is to make sure that INFORMS survives and does well. That doesn’t tell you much. The fact is, when you get into a job like this you have things that you want to do, but very often you are not able to do those things because something else comes up. You don’t know what’s going to hit you during the year. No matter what happens, though, I want to make sure that INFORMS stays on track. We have, for example, several new initiatives such as continuing education and certification, new areas where we are trying to build presence and become known as the provider of choice. Because they’re new, it’s risky business, so we have to watch them closely and make sure we make course corrections as needed.
As you noted, INFORMS has been very aggressive in launching several big initiatives over the last year or so, including CAP, continuing education and now the maturity model program, all of them aimed at the analytics movement in one form or another. How are these initiatives doing in terms of meeting goals, impact, etc.?
I think they are doing pretty well. Certification started off very well. It’s recognized as a desirable thing. We need to increase the numbers, and we expect to do that over time. Continuing education is a similar situation. We had a very high-quality start in terms of the courses we offered. The numbers are not what we would like them to be yet, but this is a start-up and few people know we provide this program. So we have to push a little bit to make more people aware of what’s available to them and help them decide to come and take advantage of it. Those two initiatives have started off well, but we need to watch them and build them because things like that don’t just blossom overnight. You have to build them up bit by bit.
INFORMS is planning to launch a maturity model program in 2014. Can you provide some insight on that program and its goal?
The goal is to give organizations the ability to apply a relatively simple questionnaire-type test for their operations and see where they are along the scale of development in analytics. It’s really a prospective tool. If you’ve got an organization and you’re not quite sure about all the details of analytics but you know you ought to be doing it, you need to assess where you are and what you need. This is designed to be a simple tool to help people do that. The idea is that INFORMS should be a source of information for organizations that want to develop themselves in terms of analytics maturity, and this program will help them understand where they are and what they need to do next. Hopefully it will also persuade them to deal with us for some of the things they need.
That sounds almost as if INFORMS is going into the consulting business.
INFORMS isn’t going into the consulting business; that’s not what we want to do. One thing that we really should be doing and that does fit our mission is we should be an information exchange, an information center. Let me give you an example. Suppose I have some difficult project and I need a certain kind of O.R. technique. How can I find somebody who is not a charlatan, who knows how to do that, whom I can hire as a consultant? INFORMS should be able to provide information about services that are available in our areas of expertise and help people get in touch with providers or consultants who offer those services.
You’ve been a long-time, active member of INFORMS and served in many leadership roles, including head of publications, treasurer and secretary leading up to president-elect. How did those roles prepare you for the presidency?
It’s helped in a number of ways. Years ago I was editor in chief of Mathematics Operation Research, so I know what it takes to edit a journal and what the problems are and what sort of issues the editors and journals face. After that I was head of publications for the two societies that preceded INFORMS where I got a feeling for how publications work at the operating level. I have also held officer positions, secretary and treasurer, and in those I got a better feeling for how INFORMS operates, its business model, how we finance things, where the problems are and how we’re organized. Doing that enabled me to meet a lot of people and to participate in a lot of things, some of which worked, some of which didn’t work, and I learned from that. So I think I have a good understanding of the way this organization functions and what you need to do to get something done and how you need to put things together to have a good chance of success.
Put into perspective how important journals are to INFORMS.
They are very, very important. They used to provide two-thirds of our revenue a few years ago. Now the fraction is down a little bit because other areas are moving up, but publications are still the major part of our revenue. We are really dependant on publications. Our journals are high in quality, they are recognized that way, and we want to keep it that way.
A new journal, Strategy Science, will soon be added to the portfolio.
I haven’t been closely involved with it, but I’ve been watching it go through the pipeline, and I’m happy to see that because it means we are moving into areas where people need information.
Meetings and conferences are obviously another very important component of INFORMS. They seem to be doing quite well.
They are doing very well. In the last three years, attendance at the analytics meeting has doubled. The attendance for several years now at annual meetings has been in the range of 4,400 to 4,500, which is wonderful. These are meetings where you can find practically every kind of information that you would possibly want. I am very happy about how the meetings are going.
The annual meeting appears to have bumped into a very high-ceiling in terms of number of attendees and a limited number of facilities that can accommodate it. Has it become too big?
Perhaps. We are reacting to that by starting a new series of smaller, special purpose meetings [such as this year’s healthcare conference and next year’s inaugural big data conference]. I think those will be very helpful. One of the areas I’m very hopeful about is healthcare. Based on the demographics in this country and elsewhere, you see aging populations and that means more healthcare demand. I think this is going to be a growth area for quite a few years.
For many years, INFORMS membership was in a slow decline. Now it appears to be heading back up. Does the INFORMS Board place an emphasis on growing membership, and what do you see in terms of membership potential?
Many people like to be members and want to be members, and we certainly welcome them to be members and we’re happy to have them. There are also many other people out there in the O.R. and analytics community who might not want to be members of INFORMS but might want to be members of our subdivisions. We are happy to have them. We are also happy to deal with people who might not want to join either one but want to take advantage of the services we offer, such as continuing education, certification and that kind of thing.
My view on membership is that it is good for some people and it’s not right for other people. We can serve both populations. Of course, we pay particular attention to our members – what they like and what they want to see at meetings and that sort of thing. I’m not greatly concerned about pushing the membership to double what it is now, for example. If we can do that, it’s wonderful, but there’s probably a natural number of people that are so interested in what we do that they want to be part of the organization. But we are also providing things that help people who aren’t members.
As editor of Analytics magazine, an outreach program aimed at non-members of INFORMS, I’m well aware of practice-oriented analysts who pick and choose the things they want out of INFORMS. They may read Analytics magazine or choose to attend the INFORMS Analytics Conference in the spring or the focused healthcare conference or the upcoming big data conference, but they don’t necessarily see the value of a full-blown INFORMS membership.
No, and that’s fine. We need to watch our organizational position carefully so that we can take care of the members with the things they want and make sure they are getting what they want to have, but we also need to price these other services in such a way that we can cover costs and provide people what they want at good value.
At this conference I’m hearing the word “analytics” more than the term “operations research,” and many members, particularly the younger ones, want to position themselves as “data scientists” rather than “operations researchers.” Is that a problem?
I think that’s a good thing. Let me explain why. Think of concentric rings. In the center, in the core, you’ve got operations research. That’s what a lot of members used to do, do now and want to keep on doing. They are in that field. That is a set of technical tools that are enormously useful in all kinds of applications.
We have, in the past, not put quite as much emphasis as we should have on the usefulness of those tools. We like to talk to each other about our theorems and our new algorithms, but we are talking to each other. There’s an old joke in the operations research community that I’m sure you have heard: People who do operation research can’t explain to their mothers what they’re doing. That needs to change.
What we do is enlarge that core by putting another ring around it, which is coupling operations research with business problems, governmental problems, other things that everybody needs to do and people know about, and that’s really what analytics is. It’s using O.R. tools to do useful things to process data, to get insight from data, to figure out what would be better for organizations to do next. We haven’t emphasized that very much in the past. Now this movement toward analytics is emphasizing it, and it’s having an effect because people are coming to appreciate INFORMS who never did before.
That’s the way I think of the movement in analytics. It’s not replacing operations research – operations research is the core, it’s going to be the core – but analytics makes it useful and makes it something that the community can relate to whereas it was tough for them to relate to operations research. The thing that we need to do that we are doing now that we didn’t do before is to get out and show people why we are relevant to them.
So the analytics “wave” that INFORMS is riding is good for the Institute and all of its members, including those who consider themselves “operations researchers”?
Absolutely, because their results, their techniques, the tools that they develop are going to be used by more people. That’s a good thing.
In your candidate position statement, you said that the INFORMS president does very little of the actual assessment work but is responsible for seeing that the work gets done well. Please elaborate.
Even though I worked in publications for INFORMS and its predecessors, I can’t go in and do a management review of the publications department and tell for sure if they are doing what they need to do. I have to know where to go to get people to help to look and see if we are doing the right things. It’s the same thing with meetings. I don’t know how to run a meeting, but there are a lot of smart people who do, and there are people who can help us assess the meetings function and see whether there are things that we can do to improve.
What I meant by that position statement is that I don’t have to be an expert on publications, but I have to make sure that publications are running the way they ought to be. I’m responsible for quality checks; making sure that everything is the way it ought to be even though I can’t do it myself.
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?
That this is a wonderful society because of its diversity. You’ve got marketing, you’ve got transportation science, you’ve got people doing optimization, you’ve got people doing organization science. Where else would you find all of these people coming and talking to each other at the same meeting? Most societies – take IEEE for example – know what they are and they do it well, but they don’t have the diversity that INFORMS has. We’ve got so many people with so many different skills, and to me that is very important because some of these extremely difficult system problems that we face require an array of different skills. That diversity of skills is what excites me about INFORMS.
We’ve talked about the many strengths of INFORMS. Publications. Meetings. Diversity. I would add volunteers to the list.
I agree. I think it’s wonderful. These people are willing to come and work hard for zero pay on committees and boards to help INFORMS. They absolutely make the place go, along with the staff. It’s been like that since INFORMS and its two predecessors began. All I can say is, in my time as president I want to make sure that we try to recognize the volunteers and help them understand that they are essential to making things run.
You’ve been an active volunteer of INFORMS and its predecessors for decades. What motivates you?
I just find it stimulating. It’s fun. There are so many good people doing wonderful things, the meetings are interesting, and it’s something I felt was worthwhile putting time into.
In your position statement you said there is “room for improvement in the way we develop volunteers.” What do you mean?
We don’t have a good system for bringing people along. Let me explain what I mean, At present, if I were a chair of a committee, for example, and some people rotated off the committee and I wanted to figure out who we should get on, I’d think, well, “Who do I know that could do that? Oh yeah, there’s Joe and there’s Mary.”
That works most of the time, but it means you are really doing random selection just based on people that somebody happens to know. It doesn’t make the young people who might be eager to get into doing something as visible as they should be, because the old guys don’t know the young people very well so they don’t think of them first. Instead, they think of people they know.
I’d like to see a system within INFORMS with a database of members who have served in various positions. So if I had a certain committee I could say, “Who has previous experience in this area, and who, if anyone, has expressed interest in this area?” and get a list of names. I could then look at the qualifications because the people that have already been serving in that area would be noted and I could get some information about how they performed. It would be much better.
Sounds like more of a systematic approach rather than an ad hoc approach.
Yes. Businesses have been doing this sort of thing routinely for many, many years. Bringing new people into the system, grooming successors, identifying young people that are sharp and are interested in helping. INFORMS needs to get them into the pipeline. Once we get them involved, then they are going get to know people and progress pretty well.
Given the current economic situation and the launch of so many new initiatives, how is INFORMS doing financially? At this conference [former INFORMS President] Terry Harrison said that the Institution is as strong financially as it’s ever been. Would you agree and elaborate?
Yes. We are currently having very substantial operating surpluses. You could say the reason we are doing well financially is that they got rid of the guy who was treasurer from 2007 to 2010. [Robinson chuckles because he happened to be the treasurer during those years.] That was not a good time to be treasurer. Seriously, we are in very good financial shape at the moment, the permanent reserve is quite high and the operating surpluses are great, so this is a time when we can afford to invest in things that we want to develop. We are investing in continuing education. We’re investing in certification and other areas like this maturity model program. You can’t invest if you don’t have money, and at the moment we must be doing something right, because we have money and we are going to try to make sure it stays that way.
Non-profits such as INFORMS can’t just hoard cash, can they? That’s not their purpose. Their purpose is to serve the members and the profession, and, as you say, expand their services and programs when they have the money and hold the fort when they don’t.
Absolutely. We do hoard cash, but we do it in a manner that our lawyers tell us is quite acceptable for a non-profit. We have a permanent reserve at the moment that covers I believe 75 percent or 80 percent of our operating expenses, and that’s pretty good.
We need that because those who are old enough to remember when ORSA just about went under realize why you need to have a reserve. Our auditors tell us that in terms of financial strength, if you look at the societies they are familiar with and rate them from one to 10 with 10 being best, in terms of financial strength, we are a 10 and we want to keep it that way. If 2008 comes back again, we want to have the financial strength to survive because we have enough cash in place.
At the moment we have in place a formal way of considering funding initiatives that was spearheaded by Don Kleinmuntz and some other folks a few years back. We’re using that to finance these new initiatives that are coming along. With the expanded view we’ve enjoyed moving into the analytics space and seeing different opportunities and new populations that we can serve, we need new programs and services, so we have to keep doing this.
Where do you see INFORMS and the O.R. profession 10 years from now? Will it even be called the “O.R.” profession, or will it evolve into “analytics”?
I would like to see us 10 years from now highly visible and highly desired in the business and governmental communities as sources of expertise, both as providers – consultants and the like – but also as generators of knowledge and expertise. These are the core people in the operations research area who are coming up with new algorithms and new ways of doing things. These have a long lead time, some longer than others. The stuff that I’ve done over my career typically had a lead-time of 30 years from the time I published it until the time it was being used broadly in the community. Some things, thank goodness, are quicker than that, but there is generally a long lead-time in this area. I would like to see us recognized as the place to go to if you need something in this broad area of how to use mathematical methods and tools to help solve problems better, which is what we do.
Peter Horner (firstname.lastname@example.org) is the editor of OR/MS Today and Analytics magazine.