The Art of Talk Design

Drawing on her extensive experience of conference speaking, Mel shared insights that are a must-see for anyone working in developer relations.


  • 5 Phases of Design thinking

    • Empathise

    • Define

    • Ideate

    • Prototype

    • Test

  • Empathise

    • Understand who our users are and what our user needs are.

  • Define

    • Observations that we’ve made so far and defining core issues

  • Ideate

    • Generating as many ideas as possible.

  • Prototype

    • Creating a scaled-down version of the end products.

  • Test

    • Getting deliberate feedback about your product and then making sure that you ask the right questions


  • Designing an effective talk, it’s the same process as designing an effective product.

  • Company launches a product after doing the research of figuring out whether it’s a product that people want, yet, it’s something that happens all the time in our talks.

  • Apply design thinking when creating a new talk?

    • Design thinking is a way of thinking deliberately about what you’re creating and constantly reacting and reflecting on that.‌

5 Phases of Design thinking

  • Empathise

  • Define

  • Ideate

  • Prototype

  • Test


  • When building a product, we first try to understand who our users are and what our user needs are.

  • It’s all about doing the research beforehand

  • Researching our users, who will be listening to it, who is the target audience?


  • Know about the events, the organizers and most importantly, the attendees.

  • What the background is, what prior knowledge they already might have.

  • Figure out what other speakers there are and also what they are talking about because that also influences people’s prior knowledge.


  • Understand what the logistical constraints are.

    • How long is your time slot?

    • What time will you be speaking at?

    • What type of stage are you speaking on?

    • Whether you’ll have access to Wi-Fi or audio?


  • Write down what you already know about the content of your talk.

    • What’s the title, the abstract, the key takeaways?


  • Analyze all the observations that we’ve made so far and define what our core issues are.

  • So what are the problems that we’re trying to address with this new product?

    • it will always be about convincing the audience of whatever message you’re trying to get across but the angle of that message can be quite different.

    • It might be about teaching a new skill, it might be to convince people to use a specific product,

    • it might be to inspire or motivate or frighten people to do or change something.

Think, Feel and Do

  • Filling up columns of as many ideas as possible with the following questions.

  • Think

    • What do you want your audience thinking about?

    • What mindset do you want to change?

    • What will the audience learn’

  • Feel

    • How do you want the audience to feel?

    • Inspired, motivated, scared?

  • Do

    • What actions do you want the audience to take?

    • What should they do after the talk?


  • About generating as many ideas as possible and then limiting and choosing.

  • Diverge

    • Try to generate as many ideas as possible while keeping the goals and the audience outcomes in the back of your mind.

    • Write down all the possible ideas that will help achieve those goals and those outcomes.

  • Converge

    • Distilling all that information into something more usable.

    • Three techniques - prioritize, filter and cluster.

    • Prioritize

      • Putting the ideas that you have in order of importance, highlighting which ideas are the most important ones.

      • Makes it easier to see which ones you might need to drop if you need to.

    • Filtering

      • Dropping off the talk, so which ones just don’t fit in, or things that aren’t as important as the rest, or that you know that you won’t get around to.

    • Clustering

      • Which ideas go naturally together?

      • Can help you identify maybe the three topics that you can cover in a 20-minute talk and maybe you can know from…you can find out from it which topics you might need to ask for a 50-minute talk.


  • Creating a scaled-down version of the end products, which you can then put in front of users and test with.

  • It’s about creating the bare minimum that will get you useful and valuable feedback without really wasting too much time.

  • Prototype your talk structure.‌

    • Before even creating any slides-- Create an outline first.

  • Beginning, middle, and end.

    • Beginning

      • Start of the talk will be in that “what is” phase.

      • The Hook

        • Grad their attention.

      • The Goal

        • Convince them why they should listen.

    • Middle

      • Everything in b/w “what is” and “what could be”.

      • The Main Ideas & Journey

        • Focus on what will bring the audience close to your goal.

    • End

      • “What could be” phase.

      • The Summary

        • Recap the main ideas

      • The Kick

        • Give them something to remember


  • The testing phase is all about getting deliberate feedback about your product and then making sure that you ask the right questions to get that feedback.

  • Prepare your feedback givers.

  • Prepare actual feedback questions.

    • What areas do you want feedback on?

    • Is it about the timing of your talk, specific sections that you don’t feel are quite right?

“Good presentations, good blog posts, good documentation, they’re all about understanding your audience and building the right message for them.”

Last updated