Blog

What the audience wants to hear vs. What I truly think

Jul 2, 2022

When selling a product, the reasons why customers decide to buy it may sometimes differ from why the inventor made it this way. Likewise, when I do research, the reasons I like a project often differ from why others find it valuable.

Here are some (simplified) true stories:

1. Question: Why did you do this research? Why is it useful?

• What they want to hear: It gives us power. It can automatically produce a program in a different language or platform, which can be faster, more secure, easier to understand and maintain, etc. There are many uses.
• What I truly think: I did it because it's cool. I initially didn't care about the use cases and had to ask others to come up with them years later.

2. Question: Why should I care about your research?

• What they want to hear: Look, there's a huge problem in how people work with software today. My research takes the first step toward solving this problem. It is promising because we have these preliminary results.
• What I truly think: I don't know. It's fine if you don't care. :)

3. Question: Why is this problem hard?

• What they want to hear: People have been attempting to solve similar problems for decades. Their solutions are unsatisfactory in several aspects. I came up with a new solution that works great in all these aspects. I saved the world.
• What I truly think: Looks easy to me. I solved it, after all.

4. Question: Why did you choose this benchmark for your experiments?

• What they want to hear: This benchmark fits our purpose. It covers many computation patterns that people care about. They also share these common properties of which my approach can take advantage.
• What I truly think: I just saw a random benchmark my friend published and started from there.

5. Question: Do you plan to commercialize your research?

• What they want to hear: Here are some challenges before commercialization. Here is how I expect to address them. If we move forward in this direction, I am hopeful we will end up with something useful in practice.
• What I truly think: I'll let other folks follow up and do the engineering necessary for commercialization. I prefer to do the first few steps of the theoretical foundations.

Thoughts

Before last year, I almost always instantly answered the "what I truly think" versions. But they often left the audience confused or unconvinced.

Then I went through the job market, revising the research statement and coming up with the first five minutes of the job talk. I realized that audience asking those questions was not interested in my motivations. Instead, they were looking for transferable knowledge to take away and perhaps apply to other scenarios. So, I had to switch to answering the "what they want to hear" versions by default.

It's a skill about communication, but not how to clarify everyone's points. Instead, it's about understanding what the audience cares about and distilling what's relevant to them. This form of communication is needed in academia to "sell" our research to reviewers, collaborators, students, etc. I am gradually learning to play this game.