32

I would like to know how the (job) market for operations research works. My dream would be to work as a freelancer in the area of math programming. I have a PHD and several years of proven experience in programming, modeling, bioinformatics, and optimization.

There are some blogs and websites by small companies or individuals working in this area, however they do not tell how they "made it".

Can one make a decent living outside the walls of a big consulting firm or software company?

hplan
  • 431
  • 4
  • 6
  • 5
    Hint: Not all of them "made it". Advertising capabilities is one thing, having enough paying customers is another. – Mark L. Stone Aug 06 '19 at 12:55
  • @MarkL.Stone: True, this is why I am asking. I do not see how one would get paid "gigs" in this field. Is it a networking thing? I am not naive: just building a website and yelling "Here I am" wont get me any engagements ;). – hplan Aug 06 '19 at 13:12
  • There is a middle ground between freelancer and big consulting company, namely, small consulting company. Are you willing to consider that too? – LarrySnyder610 Aug 06 '19 at 13:18
  • @LarrySnyder610 It depends... possibly. However I would prefer to have some say. I really like the entrepreneurial aspect. – hplan Aug 06 '19 at 13:33
  • Should the freelancing pay for all of your cost or should this be side projects you do additionally to some other kind of work? – Doegstra Aug 06 '19 at 16:02
  • Here's what I've found you need: People who know you, like you, and have money. You may find it easier to have a broader scope than math programming, such as O.R. – Mark L. Stone Aug 07 '19 at 01:48
  • @MarkL.Stone: You are right, having solvent clients should help ;). I just mentioned math programming because I find the execution central. Currently I am "selling" my solutions mostly to biologists/biochemists. They do not care about the details of the model, but are focussed on the results ;). Of course before getting there I am analyzing the requirements/problem, searching for the right approach, crafting a model, cleaning data, and then writing and applying code. Basically I am doing the full O.R. workflow from the problem to the final solution (at least I think so). – hplan Aug 07 '19 at 07:32
  • @Doegstra: Someday I hope? I am currently working a full-time job and I have family. I agree that the best start would be to do some side projects, to see if this works out at all. However, this will just be on top - which is OK for the trial but not on the long run. My goal is to do what I really love and to have more control over my time. I know, there will be deadlines, pressure, difficult clients, (almost 40, co-founder of a company).. but anyone who did some freelance work before should have experienced, that the perceived level of freedom is just different. – hplan Aug 07 '19 at 07:59

2 Answers2

31

Here's what I've seen succeed from my experience.

0) Be comfortable with OR methods. I'm assuming this is a given for anyone who finds this stack exchange.

1) Know your markets. Some aspects of business need OR more than others. Operations management has had a lot in my experience (surprise). One buddy of mine took his clients from consulting and started his own practice doing procurement and network design. For the more technically inclined, I would think the same could happen if you sniffed around for work in planning, logistics, and scheduling. As you noted consulting firms play in this space because they're great at selling themselves. You can be too. Consulting firms can't match your rate if you're independent, and you would be amazed by the number of small to mid market firms who pay people full time to solve by hand the problems I mentioned above. Get out of your comfort zone and start going to networking events and see what people react to. Also, get on sites like upwork. My other most successful friend in OR consulting brought in all his gigs that way.

2) Kill your delivery. This could mean either being comfortable with financial analysis and powerpoint for more strategic projects or being a solid software developer for those more tactical. I think you should be good enough at both of those that you can get a job doing them alone. On the more tactical side of things, this is why you see software companies working in this space too because they're awesome at making OR methods super simple to use. If you're a decent developer though, you can deliver 80% of the value of a software company's product at 20% of the time and cost (and use some finance skills to prove it). I've seen people do this by writing their math models with an ORM (i.e. TicDat or SqlAlchemy) so you quickly get something with data constraints and few changes between dev and production environments. Additionally some background in web development can go a long way here. An API or a GUI will make it as easy as possible for someone else to consume your work.

3) Don't be afraid to jump. No one I've met doing this had a "right time" to start working independently. They worked contracts at night; they used their vacation to network; they left their full time gigs not knowing what their next contract would be 6 months from then. Worst case, you fail, you're out a little money, but you've gained a ton of experience. I think there are plenty of software companies and consulting firms who would hire you just on the story that you tried this, and they would help to show upon hiring you what skills you were missing in order to succeed.

4) Connect with us. Hit up the people on the websites you've seen. Ask them how they did it. You'd be surprised how many experienced professionals would be willing to help you if you have the humility to ask.

Best of luck and I hope this helps.

Sean Kelley
  • 646
  • 7
  • 11
  • 1
    Wow, thank you for this really helpful answer. :) – hplan Aug 07 '19 at 07:08
  • 2
    The pointer to UpWork is great. I did search other portals before, but I didn't find OR-focussed professionals there (so I was wondering if these channels are relevant at all). And I agree: Having some side-projects in the field would probably be a good start to enter the market. Thanks again for your answer, these were exactly the insights I was looking for. – hplan Aug 07 '19 at 07:38
11

I like Sean Kelly’s response. I’ll just add a few points to this.

Over my career, I’ve worked in small to large firms, always close to OR. And, I’ve run into quite a few people who have been successful at being a freelance math modeler—either I worked with them, they brought me into projects, or I brought them into projects.

Here are a few other observations of different ways of making it:

  • Some have used their experience in big consulting firms to build a network and build the skills to go out on their own. So, you could go to a big firm with the goal of figuring it out so you can go out on your own.
  • You have to be good at sales—getting projects to actually sign can be time consuming and harder than you think.
  • Related, you need to be prepared to weather the time, and more important, the lack of cash coming in, between projects. This time can seem memoryless when you are in it, and endlessly frustrating waiting for prospects to move forward or surface. Remember, when you are really busy with a project, your pipeline and network will grow cold.
  • Make sure you are ready for the other aspects of being a freelancer—collecting money for the work you do (be ready for some firms to be super slow in paying) and being able to manage expectations and tricky situations (like a client who claims you didn’t actually do what they wanted and isn’t paying)
  • Don’t put pressure on yourself to make it an all-or-nothing type of career choice. I’ve seen people who do this jump between short-term projects, long-term projects (that look almost like a job), and even taking employment with firms.
Michael Watson
  • 841
  • 5
  • 11