How to Choose a Conference Talk Topic

You’re not Beyoncé, but you should treat your talk deserves a tour

I’m a Beyoncé fan. I’m not going to argue with you online, but I’m going to every concert.

When I was younger, my sisters and I would dance to “Ego” while my mom was out grocery shopping. The second we heard the door open, we’d scramble to turn off on-demand music videos on TV and pretend we were studying. (My middle sister now claims she never wanted to dance with us, but I’m a bossy big sister, so she had no choice).

My husband is also a huge Beyoncé fan. Since we’ve had enough money and a license, he’s brought me to all of her concerts.

I’m currently reading Matriarch, her mom’s memoir. Beyoncé is a source of inspiration.

This Memorial Day weekend, I went to the Cowboy Carter concert. It was a big deal because I was away from my baby for 12 hours.

I haven’t been away from her for five months. She’s five months old right now.

I drove to a different state. I haven’t driven since October 2024. You do the math.

The drive felt like therapy: clear skies, fresh air, and Beyoncé on repeat.

I kept thinking about the clips I’d seen online. Beyoncé floating in a car above the audience singing “16 Carriages.” Sitting beside Rumi and Blue while singing “Protector.” I had already seen these moments dozens of times, but I still felt excited to witness them live.

Then, I had an epiphany: we’re not Beyoncé, but conference talks are our performances.

We go on tour with our talks, delivering them city after city, sometimes tweaking the format for different audiences. People tweet highlights and spoilers, but others still want the live experience.

Just like artists carefully curate their setlists, we have to be intentional with our talk topics. Getting accepted is only the first step. Once you’re on stage, you want the talk to matter.

You’ve traveled this far. You’ve missed time with your family. Your company spent real money to get you there. You want it to be worth it.

In this blog post, I’ll share how I choose topics that respect my time and the audience’s. I’m not writing from a place of perfection. I’ve had talks that earned standing ovations and others that made me feel embarrassed people showed up. The failures taught me the most about what makes a topic resonate.

Framework

I usually have a CFP tab open for days. I always wait until the last minute—11:59 p.m.—to submit a rushed talk I’m not even sure is good. I just want to get in.

The best rule of thumb, which I don’t always follow, is to map out talk ideas at the end of the year. Decide which 2–3 talks you’ll rotate across conferences. Think of it like a tour.

But more importantly, I’ve started reframing the question. Instead of asking What should I talk about? I ask What intersection of my goals, passions, and the moment we’re in will create the most impact?

Here’s the framework I use now:

My Goal

My goal isn’t “give a good talk.” Let’s be honest, I’m a developer advocate. I’m paid to create awareness and adoption.

Your real goal might be:

• Get engineers to try a new SDK
• Change how teams approach code reviews
• Build awareness around your company’s approach to AI safety
• Establish yourself as a thought leader

How to Find Your Goal

Ask yourself:

• Why am I submitting this talk?
• Is my company paying me to be there? What outcome are they hoping for?
• Am I building my personal brand? What do I want to be known for?

My example goal: I want more people to know that Goose, an open source AI agent exists and feel curious enough to try it. Stretch goal: They share about the tool or talk on social media.

✍🏾 Your turn: Write your goal down.

Everything Everywhere All at Once

Once I write down my goal, I feel the urge to cover everything.

If I want people to try Goose, I start thinking I need to explain how to install it, demo every feature, share its history, prove it’s useful for devs and general-purpose use, and show how to contribute. But most talks are 20–30 minutes. I don’t have time to do it all.

And then I want it to work for everyone: students, junior developers, seniors, staff+, managers, and executives.

But as I’ve learned over time, this is the easiest way to lose people. Juniors get overwhelmed. Seniors tune out. Execs don’t see the point.

When a talk is for everyone, it’s for no one.

Instead, I pick one slice. Maybe how I use Goose day-to-day. Maybe how I contributed a feature. I focus on a clear audience, like mid-to-senior engineers. That specificity brings clarity.

My example focus: Unique features of Goose that would intrigue mid-level and senior engineers.

✍🏾 Your turn: Write your focus and audience.

Passion

If you don’t care about your talk, your audience won’t either.

When I say passion, I don’t mean yelling and sweating like a preacher on Easter Sunday. I mean the kind of thing you could talk about with a friend for hours. Something that lights you up or makes your brow furrow. It doesn’t have to be technical.

Yes, your company may be paying you to speak. Yes, this may help your brand. But why does this topic matter to you?

My talks usually come from two places:

• Excitement: “I found something amazing. I need people to know.”
• Frustration: “I keep seeing the same mistake, and it needs to be addressed.”

My talks have included Beyoncé, Ratatouille, MySpace, Mean Girls, Guyana (where my parents are from), and Antigua (where I was born). I like blending culture, life, and tech.

Even during maternity leave, I couldn’t sit still. Winter felt gloomy and repetitive. So while nursing my daughter, I started contributing to open source just to stay mentally engaged. I got curious about AI agents, browser APIs, and the emotional stagnancy of parental leave.

I turned that experience into a talk. Someone told me, “That was good storytelling.” It wasn’t flashy. It was personal.

Speakers Who Model Passion

• Kelsey Hightower – At RenderATL, he took us from his emotional journey of tech support technician in Atlanta to Kubernetes icon. It felt like I watched a movie.
• Angie Jones brings personality and clarity. I felt nostalgia while listening to her talk about mashups, MCPs, and AI agents even though that was my first time hearing of mashups.

My example passion: Rediscovering the joy of building through vibe coding.

✍🏾 Your turn: Write a speaker whose passion moved you.
✍🏾 Your turn: Write down a few things you’re passionate about.

Trends

It’s not your job to start new conversations. It’s your job to meet your audience where they already are.

Look at:

• Tech chatter: What are people debating on Twitter, Bluesky, or Discord?
• Company context: What are people already curious about?
• Timing: What cultural or tech shifts open the door to your topic?

Sometimes the „trend“ is the tool or concept itself. For example, when GitHub Copilot was trending when I gave the talk Level Up with GitHub Copilot. People already had strong feelings about AI coding assistance. I didn’t need to add anything extra; I just needed to meet people where their curiosity and concerns already were.

My example trend: „Vibe coding safely.“ People are excited about this new workflow, but also nervous about doing it responsibly.

✍🏾 Your turn: What trends are you seeing in the dev community?

Titles

You’ve used the framework. Now you need a title that stands out during CFP reviews, hints at your unique angle, and sets expectations.

Let’s rewind:

• My goal: Make Goose known and spark curiosity
• My focus: Mid-to-senior engineers intrigued by Goose’s features
• My passion: Rediscovering the joy of building through vibe coding
• My trend: Vibe coding safely

How I Brainstorm Titles

First, I get into flow state:showering, lifting weights, pushing my daughter around the neighborhood in a stroller and recording my stream-of-consciousness in a voice memo. Then I turn to AI tools as a sounding board.

I’ll ask ChatGPT, Claude, or Goose:

I want to do a talk called “How to Vibe Code Responsibly,” but I want MCPs in the title too.

AI suggests:

• How to Vibe Code Responsibly (With a Little Help from MCPs)
• MCPs and the Art of Vibing Code Responsibly
• How to Vibe Code Responsibly Using MCPs, Not Magic
• Vibe Check Your Code: Responsible Automation with MCPs
• Coding on Vibes? Let MCPs Keep You Grounded

I’ll respond with something like:

I like this, I think.
How to Vibe Code Responsibly (With a Little Help from MCPs)
and this: MCPs and the Art of Vibing Code Responsibly

I’ll follow up with

I want this to show off how Goose can help with responsible vibe coding. Here’s my outline: [insert outline]. What do you think my three take aways are?

AI’s response

Vibe coding is powerful—but dangerous without structure. You can build entire apps with a single prompt, but without guardrails, you risk exposing secrets, breaking workflows, or losing work.
MCP gives agents the structure and context they need to interact safely and effectively with real tools like GitHub or Postgres.
Use tools like .gooseignore, /plan, approval modes, and version control to stay in flow while protecting your project, team, and future self.

My example title: How to Vibe Code Responsibly (With a Little Help from MCPs)

Reality Is Messy

My framework is for the perfect, ideal situations. But sometimes you come up with an amazing talk that you want to go on a tour with and hits with one conference, but flops at another. Don’t let that get you down. Evaluate and rotate your talks as needed.

You’re an expert

Public speaking used to scare me. My voice cracked. My throat tightened. I cried from frustration on stage. Medical professionals call this globus pharyngeus aka you’re super nervous.

But over time I realized:

I don’t need to pretend to be an expert. I am the expert because I’m telling a story only I can tell. Your best performance can start with a thoughtful idea.

How do you come up with talk topics?