What has giving tech talks on dataviz taught me

I recently gave two talks about visualization, one is an intro to dataviz talk+ R workshop for Women Who Code, and another is at a tech meetup talking about my project 100daysofviz.
As a not-so-talkative introvert, I took up the speaking opportunities because public speaking is part of storytelling and good data people need to be superb communicators.
Here are a few things I learnt from giving public talks for the first-time:
Be well-prepared, no matter the event is big or small
And that preparation = research + rehearsal.
Before the dataviz talk, I searched for similar talks, sought advice from interesting dataviz people, read a few books (Semiology of Graphics, The Visual Display of Quantitative Information and Beautiful Evidence), watched all openVis videos (I do want to attend a cool viz conference one day), and also went through Pixar’s Art of storytelling (which is great but not directly applicable to data projects).
While research is vital, things never turn out as smoothly as expected as I developed a fever a few days before the talk, and was not sure I could conduct the talk and ensuing coding workshop coherently. But it did turn out ok as I had the slides and materials ready and eventually get rid of the flu.
Talking about personal project was harder for me. Up till day 40 I told less than 5 people about it. Then I realised the very goals of this project is self expression (and hopefully create kickass visualization). And that vocal expression shouldn’t be harder than written expression.
After the talks, what I learnt is:
Talk it easy
To relax is possibly the most important thing in speaking. I am lucky to have not encountered any stage fright yet, possibly not having constant eye contact with the audience is the reason for it (I thought I did, but I was actually glued to the slides part of the time).
Give it a structure
In terms of what goes into a talk, preparation and improvisation equally matters. Prepare the beginning (the hook) and ending (the takeaway) is probably necessary, as they hold the overall structure, but a little free flow in the middle should be fine.
Another trick on the structure as recommended by seasoned speaker Carmine Gallo who studied 400 TED talks is this formula:
65% pathos (narrative) + 25% logos (data) + 10% ethos (street cred)
Tech talks probably will deviate from the percentage, the general idea is that to give a talk is to tell a story.
Q&A is the best part
After a few round of Q&A, I realised some people are just genuinely good at asking questions, which is an art in itself. Other than that there was a lot of interest on the logistics of this 100day project, what triggered me to plunge into it and how I carry it out. I actually feel Q&A can be fun as it is interactive and smashes the invisible wall between the speakers and the listeners.
While the two talks went well (for a speaking novice), I don’t intend to talk again in near future as it’s fairly time-consuming to prepare when I have a lot of explorative learning planned. It did teach me a thing or two about communication and about myself though.
This is #day43 of my #100dayprojects on data science and visual storytelling. Thanks for reading. Suggestions of new topics and feedbacks are always welcomed.
