I was invited to one of the biggest QA events (expected more than 400+ professionals) in my country with topic about performance testing.

The information I would like to cover is:

  • Install two docker images (Grafana and InfluxDB) and configure them to communicate to each other;
  • Configure both of them (Grafana needs some time for setting up, InfluxDB will be quick);
  • Create a jMeter test (with all the explanations of how and why) and connect it to InfluxDB;
  • Configure Grafana to process and visualize the information;
  • Rework the test case to generate dashboard report (new in version 4.0).

I have 30 minutes in total with following scheme - 30 minutes without questions from the public, 25 minutes with 5 minutes questions from the public and 20 minutes with 10 minutes questions.

Public will be very heterogeneous containing junior/mid/senior/TL QA's and others (as DevOps, HR's, etc).

The problem

As you can see the huge amount of information that I need to pass to the public and the work I need to do.

Possible solutions:

  • Create presentation containing screenshots of all steps that I do previously, no demo;
  • Same as previous, but run the tests live and analyze the results. Everything will be ready in advance;
  • Do a live demo.

What I'll not do

I went to enough events to know that reading from the slides, stuttering or uber boring academic mumbling type of lectures are the thing I would not like to give the persons who come to listen me. In most of my previous lectures I do it with few slides and mostly freestyle, but this one is different.

Do you ever had such an experience and what would you prefer if you were in my place?


I have previous experience of lecturing (including academic) and I have good background of the stuff I would present.

  • 2
    I'm voting to close this question as off-topic because it's not about the workplace? it's also really really specific to a very particular problem. – bharal Mar 31 '18 at 17:32
  • From your 3 options my vote goes to 2nd: prepared in advance and run only the interesting part live. In your description you concentrate on What and How. Please add the Why aspect. Why are you doing this? What is your goal? How would you define a success? – PM 77-1 Mar 31 '18 at 17:58
  • @bharal thanks for the comment. I used the workplace, because I'll represent the company I work in. If this is the incorrect place where I should've posed it? – Nedko Hristov Mar 31 '18 at 18:02
  • @Nedko quora, definitely. Probably reddit too. Just because it's a work event doesn't mean it's suitable here, eg "I have trouble sleeping when flying, next week work are sending me on a long trip, how can I sleep" is similarly off topic, although, like your question, deserving an answer – bharal Mar 31 '18 at 18:23
  • So feel free to ignore me if you think I have no idea what I'm talking about, but generally a presentation should be more about showcasing your results and a high-level description of the approach you used to get there, not something that intends to go into enough detail to allow anyone watching to reproduce what you did (but of course you can point people to more exact instructions so they can do that offline). The fact that you have 30 minutes in total seems to reinforce this idea. – Dukeling Mar 31 '18 at 20:20

Having done this type of thing with large and small audiences, I would say the most important thing is to keep it simple. Start with some background and context, tell a compelling narrative about what you're doing and during the course of the talk communicate ONE maybe TWO interesting or novel ideas. That's the maximum a heterogeneous audience can handle.

It is really useful to look at other talks (both successful and unsuccessful) to see what might work for you. The AWS re:Invent series are on Youtube and are pretty good for a WIDE variety of examples (NOT the keynotes-- those are too polished for a technical talk) but rather the smaller talks that are focused on more narrow topics.

If you're really good, you will deliberately leave out large pieces of detail but be ready to answer them, as needed, when questions arrive. This way you can prepare in advance a large amount of information but only go into specific details when there is enough interest to prompt a question.

Live demos are simultaneously boring and risky. They're boring unless you're showing something absolutely new and exciting. No one wants to just look at stuff they look at all day long at work, so it has to be really new if you're going to demo in front of audience. They're ALSO risky because if you mess up, it strongly detracts from your message. It goes without saying that anything that depends on a network in a room full of techies with laptops and Iphones is asking for trouble.


Well, I would wonder why i'm giving the talk. What's the point - is it to show how to setup the thing, or why someone would want to setup the thing?

When I've gone to talks, too much technical detail is boring and pointless - I didn't come with my notepad so "make sure you set parameter x to 12" is going to disappear from memory pretty quick. I prefer "here's a neat solution" and then showing what it can solve.

Thus i'd prefer you show the completed setup, and provide a link to a youtube video about how the heck you setup the thing. With the completed setup you can show why anyone wants to do this - ie what is the business problem it solves?


As with any presentation, you need to practice, practice, practice until you nail it in terms of content, depth, and timing.

If you can’t present your content adequately within the time constraints, then you either have too much content or not enough time.

If you can’t change the time constraint, then all you can do is cut the content.

Not the answer you're looking for? Browse other questions tagged or ask your own question.