How to structure presentation to avoid getting questions that will be answered later in the presentation?

Question

In the past when I've given presentations at my lab and conferences, I've not had to deal with this problem. But at my new institution I'm constantly dealing with the following phenomena.

I'm constantly being asked questions that will be answered later in the slides.

First try, I tried top down approach. So I started with the overall problem. I work in AI, so in my particular situation, I start off by saying this is the problem we're trying to solve, here's the overall architecture and now let's analyze the individual components. But then, I'm constantly bombarded with, well what about this component, what about this metric, can you go deeper into just that one component - I don't understand the mechanics behind it, etc. And all these questions will be answered later on. But no one seems to wait to see what the next slides are going to be.

Second try, I tried bottom up approach. "Here are all the individual tools you'll need to understand the overall architecture. Here are the metrics, here are individual mathematical concepts, etc. Now, let's see how these are incorporated into a larger framework." But before I can explain what the larger frameworks are, at every slide, they're constantly asking, wait, why is this being used, why are you talking about this, etc etc.

I understand questions arise and I can simply say I'll cover it in the next few slides. But there are sooo many questions being constantly asked and it completely throws off the presentation because constant tangents are being drawn and one person's question confuses another person. This isn't a classroom setting, it's more of a lab setting.

How can I get around this? The only solution I can think of is to have a little box on each slide, right about the page number that has the title of the next page.

Any advice on dealing with this would be very helpful!

Update: After listening to everyone's suggestion, the following tips really helped solve this problem.

1. Preparing slides to control the conversation. Focusing on only what I prepared instead of making all encompassing slides. For example, if I want to convey the training procedure of a neural network, then only have details regarding input, output, architecture(not every detail, just overview), and the loss functions. Stating "I used equation 'x' because it does will in situations 'a' and 'b'," instead of pasting the actual equations and inciting more questions for understanding every detail of the equation (as was happening previously). But I keep more detailed notes in the appendix slides in case anyone has a very specific question.
2. If someone does have a very specific question, then I can refer them to the appendix of these slides and say "Thank you, that's a great question. I have lots of information regarding that in the appendix of these slides. But due to time constraints, let's take that conversation offline."
3. Sending the audience the slides and notes before your talk so that people can review the slides before hand and better prepare their thoughts before asking random left-field questions during the actual presentation. This doesn't work in conferences, but definitely helps in lab settings.
4. When stating a key concept in a slide that hasn't been addressed previously, I have a small textbox that says "will be covered in upcoming slides" in red lettering. This seemed to curtail a lot of questions as the audience no longer feels anxiety over not understanding something on the slide.
5. Right after the title slide, I had a "goals of today's talk" slide in which I stated "goal is not to explain every detail of a particular technique. Rather present overview and my adaptation of it."

Disclaimer: the above suggestions work great if you're presenting at an internal team conference amongst peers. For example, an R&D lab at a tech company (my situation) or in an academic lab setting. Doesn't go great in a conference, but the audience are generally kinder in such public avenues.

Answer 1

I'm constantly being asked questions that will be answered later in the slides.

You have to understand that this is a great problem to have. There is nothing worse than giving a presentation and facing a bunch of blank stares. Questions mean that

1. People actually understand what you're talking about.
2. They care enough about your idea to actually ask.
3. The problem you're explaining is not trivial.
4. You actually have an answer to the question - later in the slides!

You mention both conference talks and lab talks. In the first case you need to be a bit more prompt (always respect the time assigned to you in a conference, you're not alone and you'll put everyone in an awkward position if you go over time): just say "this is a great question, can we wait for the next slides?" or "can we take it offline?". I have yet to have seen anyone respond to this by insisting that you answer them right now.

But there are sooo many questions being constantly asked and it completely throws off the presentation because constant tangents are being drawn and one person's question confuses another person.

This is less a problem of structure than a problem of you not controlling the discussion. If there are many questions and people start discussing amongst themselves you need to take command of the room - say, "it's great that we're having this discussion, but I'd really like to move forward!", or something to that effect.

I believe that neither of your approaches (if I understand them correctly) is as effective as it can be. In conference presentations you need to focus on one big idea and explain it well - the objective of the presentation is to get people to read (and cite) your amazing new paper. Avoid giving broad overviews, and getting into every little detail. The audience needs to know why they should spend time reading your paper, and that's it.

In lab presentations make sure that the group knows what the big plan for today is. If there is a constructive discussion that's great. Tell them in the beginning the big points and then try to cover everything else. Give references to stuff you didn't cover so that the group knows where to look for things that weren't covered in the presentation.

Good luck!

Answer 2

Make it clear at the start that you will take questions at the end.

Then, when someone interrupts with a question, don't answer, but say: "please keep that question for the end" and continue.

If you stick to it then it will work, if you answer just one question you have failed...

Answer 3

I completely agree with the answer given by Spark. It is as much of a luxury problem as it ever gets in academia (maybe with the exception of having too much grant money and too few projects/staff).

To add on to that answer, I would like to give a couple of tips/insights on presentation techniques which may help you see this from a different angle.

First off, remember it all makes sense for you and for you alone. Often enough, connections and relations that may seem obvious to you will not be as obvious to someone else who has been thinking about other problems up until (and possibly even during your talk). So you need to take your audience by the hand and guide them through what you want to say. (Also related, but slightly off-topic, there is a saying in literature "The author is dead" meaning when you are reading something you cannot ask questions since the author is not there. However, during a talk you can very much ask questions about bits and pieces that do not make sense for you. Some people do that, and at time over zealously)

So, what can you do? I for one appreciate the introspection you seem to show and that you have tried different approaches. Either approach is fine, I'd say, but as you have commented neither is a solution alone.

I would suggest having some "meta-talk" within your talk. Overview/Agenda slides are particularly good for this purpose. You can tell people what you intend to go through/cover in your talk, how the concepts relate to each other, without going into any detail, in the beginning of your talk. That way you can prepare them for what is to come. This is also a good time to point out that some details will be revealed later, perhaps after you introduce the concepts but not their relationships to each other. This is a highly effective way to keep people interested, actually.

If questions arise, then as Spark also pointed out, you can quickly swat them away by saying something like "I am glad you caught that, let me come back to that in a minute". That way you acknowledge the wisdom and relevance of the question, without getting derailed. It takes some practice to keep your composure and stay confident, when your flow is interrupted, especially it happens multiple times.

I am not sure if this is an issue for you, but by getting frustrated or confused you might leave the impression that you are not actually sure about what you are saying and some people (especially on a conference setting) might see that as a sign of weakness and keep prying into it.

Finally, if despite your best efforts people keep interrupting you in your own lab meetings, you can bring it up with whoever is chairing these meetings and tell them that maybe questions should be kept for the end, or wherever the belong, in order to not hurt the flow of the talk. I cannot imagine an environment where people don't care about the fact that they are hurting a colleague's flow.

After all what is the point of going to a talk if you don't allow the speaker to speak?

Answer 4

Structure is everything!

The first rule to managing a presentation effectively is to provide an overview before going into details. This follows the classic 3 step process of

1. Tell them what your going to tell them in the presentation,
2. Provide the main presentation,
3. Tell them what you told them.

You can spend a short amount of time verbalizing step 1 while providing a succinct written agenda. The agenda will provide everyone with anchor points that you can refer to when questions come up during step 2. (It's often helpful if the agenda is always visible to the audience, either in their own copy or as a sidebar on the slides. You can get creative with this.)

If you're at a conference and time is an issue, you can safely skip step 3.

The point to doing this is that when a question of the type you describe comes up, you can plausibly tell the questioner that you will get to the subject of that question at a specific point later in the presentation (tell them where that lies in the agenda) and ask them if they wouldn't mind waiting until then. This is usually sufficient to get them to hold their questions without objection, and avoids the unwanted tangents that you refer to in your question.

Answer 5

When do you want to answer questions? If you want to answer questions at the end then say this at the beginning and stick to it. If you want questions asked throughout at specific moments, then perhaps have a question slide after each section and inform the audience that you will be doing this. The audience is along for the ride here and you can fill them in on the plan so they get the optimal experience.

Answer 6

One more approach you could try is a chronological presentation: describe your actual process of solving the problem. Presumably you did not work on entirely separated components and combine them only at the end (bottom-up), nor did you have such a good overview at the beginning that you could break it up (top-down).

Stepping through your progress should limit the number of "why don't you do it this way" questions, because you can easily give design rationale. Because you are building up the system as you speak, the components are directly connected to each other (unlike bottom-up), so it's clear how they relate to each other. And because the audience has an understanding of how far you are with solving the problem, they should understand that some things will only come later.

Of course, a top-down overview at the beginning and/or the end can still be very helpful.

Answer 7

If you have the opportunity during preparation, one technique that helps with this is to present a draft of your talk to a test audience and get feedback about what things are confusing or seem ungrounded at the time you mention them. As mentioned in other answers, the context of each piece of information makes sense to you, since you already have the full picture, but by getting the perspective of someone who isn't already familiar with the topic, you can get a better sense of how to sequentially build on the information you're presenting.

Answer 8

I have a love-hate relationship with agendas in presentations, for that exact reason.

I do not like to describe what I will be talking about because, well, I will be talking about these things anyway.

On the other hand your question is exactly why I use them: so that people do not wonder whether I will mention wazii and wazaa. A well built agenda (not like the example above) can help them to cool down the excitement and wait for slide 54 where it is discussed.