(Also see my notes on writing a technical paper.)
A poster session at a conference gives you an opportunity to present preliminary work (that is not ready for a published paper or an official talk slot) in a venue that permits one-on-one (or one-on-few) interaction with interested conference-goers. The greatest value of the poster is when you are standing in front of it, but during other times, it can still convey information to passers-by.
This document gives a few pieces of advice about preparing and presenting a poster. They are also relevant to technical demonstrations, which have many similarities.
Use the visual space effectively. Use graphics. The poster should draw people to learn more, especially from you but also from the poster itself.
Do not design your poster as a set of slides, arranged in a grid on the background. This is very easy to do, so it takes little of your time. However, it discourages foot traffic. The format is unattractive, viewers have difficulty grasping the essential idea and the connections among the parts, and it is clear that you didn't care enough about the topic to make a good poster.
As a corollary, print your poster on a single sheet of paper rather than on multiple pieces of paper. Use of a large-scale plotter is expensive, but your trip to the conference was even more expensive (and so was your salary while performing the research), so it's worth it to make your impact as large as possible. If you don't want to travel with the poster, you can have it printed in the conference city, if you have time to travel to a print shop there.
Bring your own pushpins, thumbtacks, or binder clips. The organizers should provide these, but sometimes they are nowhere to be found.
Before starting to stand at your poster, supply yourself with food and (especially) drink. You don't want to leave your poster unattended, and if you are lucky there won't be any dead times during which you can leave anyway.
Give the audience a handout to take with them, which will help them to remember your research. One good choice is a single sheet of paper that reproduces your poster on one side, and has a one-page description of your idea/technique/results on the other side. Don't forget to prominently include your contact information, and URLs for downloads or other information. Bring more of these than you think you will need; you may be surprised how fast they disappear.
You can put the handouts in a manila envelope and attach this to your poster (the tape, binder clips, etc. will need to be strong!). Write "take one" in large letters with a magic marker. Also, cut away part of the manila envelope (make a "scoop neck" opening for the envelope) so that people can more easily see the handout and remove it; this makes a difference in how many people will take one.
You may wish to use a second manila envelope that you prominently mark as "leave business cards, comments, and questions", so that when you are not at your poster, you can still make contact with people who are interested in your work.
Prepare (and practice!) a one-minute “elevator speech” or "stump speech" that conveys the key ideas of your work: what the problem is, how you solved it, and what the results are. Do not let yourself go over one minute. This speech will help to orient people when they walk up, and permits them to ask questions of interest to them, which starts your conversation with them.
While you are in a conversation with one person, other people will wander up. At this point, you are on the horns of a dilemma. You want to engage the newcomers, so you shouldn't get too deeply embroiled in a detailed technical conversation with one person. However, the newcomers may be interested in the same questions, and you can often draw them into the conversation or give a 10-second overview of the idea you are discussing. (A glance at your poster should give the context in any event.) In some cases, it may be appropriate to make an appointment to talk more with the interested person at another time.)
Back to Advice compiled by Michael Ernst.Michael Ernst