Pitching the right work to the right audience

15 December 2017

For the entire month of October I got to remember what the sun was. I travelled back to Australia from the UK to visit my family and attend a wedding of a very good friend. But while I was there I spent some time at my old university, the University of New South Wales (UNSW). This gave me to opportunity to speak to my old PhD supervisor and a professor I briefly worked for. In addition to this stay at UNSW, my wife had the wonderful idea of travelling to Melbourne to visit the universities there, specifically Monash University. So all of a sudden I was on tour in Australia.

Now, the groups that I was visiting in Sydney and Melbourne are quite different. At UNSW, I was giving a seminar to the applied mathematics department and at Monash University I was visiting the School of Information Technology. So just from the names of the department, it could be assumed that these two groups would be interested in different types of work. So I decided that I would tailor my work to each one.

I said three seminars at the beginning, but only spoke about two. The final one was back at Lancaster University for the Data Science Institute. This group seemed to fit nicely between the applied mathematics and the information technology audiences that I previously presented to. So there I needed a third seminar.

Why present different work?

I believe that in giving a seminar, it is not about you but about the people you are talking to. You are the entertainment for an hour and you should try to make than hour as enjoyable as possible. If you walk into the seminar not caring about the people you are presenting to, then you will completely loose everyone in the room very quickly. They may still be there physically, but you can be sure that they have drifted off to think about the mountain of work that they need to complete in the next week.

So it is my ambition to be able to leave a seminar with people finding something interesting about the work I am presenting. This can be difficult, and it can also come down to your presentation skills. However, the more interested your audience is, the easier it is to present to them.

What did I present?

So the first seminar was for the applied mathematics department at UNSW. This seminar was primarily designed to let my old PhD and postdoc supervisors know what I have been working on recently. So the work here was to target a very applied audience. I was fortunate to work with Benjamin Müller, Ambros Gleixner and João Pedro Pedroso on a very applied optimisation problem, namely the recursive ring packing problem. Now, I put Benny at the start of the author list there because he was instrumental in completing the implementation and writing for this work. The recursive ring packing problem may not sound very applied, but when you know that it comes from the packing of tubes into containers for transportation, then it really fits with the applied mathematics audience.

My next seminar was pitched at a information technology audience. This one was not too difficult because the group has a high level of expertise in combinatorial optimisation and constraint programming. So I could really have spoken about anything related to my work to this group. But I wanted to really appeal to their expertise. Currently, I am working on a general implementation of Benders' decomposition for SCIP. So it gave me a good opportunity I talk about the recent developments. At present I have been focused on the use of heuristics for Benders' decomposition, in particular large neighbourhood search heuristics. This work turned out to be very computationally heavy, and not as much mathematics, so I felt that this was a good fit.

Finally, upon my return back in "sunny" Lancaster, I was presenting to the Data Science Institute. This is a multi-disciplinary group, so this was probably the hardest one to pitch. In this case, you needed to have something interesting, but not too technical that you lose most of your audience. Luckily, when I was working with John Murray I was involved in a project that could easily fit within data science (even though I don't really know what this means). This project employed mathematical programming in attempt to find potential positions of the HIV env sequence that could be used for a vaccine. It seems pretty cool, I am using mathematical programming to help in the development of a HIV vaccine. This talk had many parts to it, mathematical programming, heuristics, networks and the interesting application of HIV sequence analysis. So I felt that this really fitted with a multi-disciplinary audience.


Now, I can't tell you if pitching to the different audiences worked, as there is not real control for this experiment. However, I was very happy to have selected my current work to present to the research group at Monash. This was probably the best decision I made in regards to the seminar choices. There was many positives from selecting this work for Monash. The first was that people there were working on similar benchmark instances, so that was a big win. Also, the group was interested in the use of large neighbourhood heuristics for other problems. Overall, I just felt like it was well received.

In regards to UNSW and the Data Science Institute. At UNSW, it was just very relaxed, so I could have talked about anything and it would have worked. I just took the opportunity to talk about a piece of work that I find particularly interesting. For the Data Science Institute, it turns out that there is a member that works on similar bioinformatics problems. So it was a really good pick of seminar topics in order to connect with a potential new colleague.

In any case, selecting the right seminar for the right audience can only help you. It is easier for a group to remember you if they were interested in what you were saying. Also, it is more fun for you giving a seminar if your audience is more engaged.

© 2018 Stephen J Maher
Template design by Andreas Viklund with modifications by Stephen J Maher.
This page was last updated Tuesday, 23 January 2018.