Post

Phew! Having reviewed 163 CFP submissions for @localfirstconf.com over the last week, I have some thoughts Here's some tips for an effective talk proposal!

1. Pitch me, get me excited Make your proposal a short, exciting teaser for your talk. Keep backstory minimal. Highlight problems that speak to your audience and the unique insights you've gained through hard-won expertise.

❌ Bad "In this talk I will discuss Project Giraffe, which is a prototype I made some years ago but never released. Maybe I will publish the source code before I give the talk. I had some different ideas about data migrations. I will try to do a demo if I can get it running again."

✅ Good "Migrations in distributed systems can be a nail-biting adventure in avoiding data corruption. I led the team which handles this at BigScaleCo, where we found that standard transactional migrations break down. Instead ...

... we use rolling migrations with a library we built for the purpose (now open source), data lenses for realtime data, and a live regression monitor to catch corruption before it multiplies. Plus: why we've counterintuitively found that Friday afternoon is the best time to run migrations!"

2. Tell me who you are Ideally two links: your bio website that lists the major projects of your career, and a link to the company or project you're representing. Don't drown me in numerous links, or make me hunt around to figure out what your career has been about so far.

3. Show proof you are good at presenting Link to a recording of a past talk you've given. If that's not available, a five-minute Loom/CleanShotX "talk to the camera" screen recording demo or presentation of your project is fine.

4. Make sure you have real expertise on the subject We sometimes get submissions for unreleased prototypes or quick weekend side projects. These are rarely compelling. Instead we want to see that you have hard-won expertise and deep insights to share with the audience.

(Side projects are fine! But show that you've worked on it a long time and have some genuinely interesting insights from it.)

5. Don't submit a lot of proposals It's fine if you want to submit two different ideas because you're not sure which the organizers will like better. I recommend against submitting more than two. Even better, pick your best one and send only that.