On Friday, December 11, seven TKS students gave talks on topics ranging from tackling climate change with 3D printed meat, to designing a product to enforce social distancing, to the value of self-restraint from pleasure. These were high-standards talks, keeping the audience engaged and learning throughout the night. You’d expect nothing less from TKS kids.
These talks didn’t occur on a stage like TED or Evoke, though, or their virtual equivalents. They were part of the first-ever round of TKS Community Talks, organized by me with the help of another Activate student, delivered internally to the TKS Community over Zoom. More than fifty people showed up to watch, and a few recordings have already surpassed 100 views on YouTube. Eighteen people have filled out an interest form for speaking in the next round.
All in all, this first round of TKS Community Talks was a resounding success. Respondents to a post-event survey marked an average score of 8.8/10 to the question “How much did you enjoy this event?”, with 10 being “Favorite TKS event” and 1 being “Not at all”. A rating on professionalism averaged 8.67/10, and value 8.33/10. 100% of respondents said that they would attend another session. Three speakers asked to help organize the next round, and enthusiasm has been high throughout the community.
In this post, I’ll give an overview of how this first round came together; what went well, and what didn’t; key learnings to apply to future rounds; and initial thoughts on the future of TKS Community Talks.
There are three big objectives of TKS Community Talks:
Inspiration for the first objective came after one of the first Innovate sessions on presentations. Being able to give high-standards presentations is an invaluable skill, whether you’re pitching to an advisor, investor, or audience of 10,000. Before that session, though, most students’ only practice with presentations was in school, where the goal was far from giving a good presentation. The Innovate presentation session taught students to unlearn school-based ideas and learn what a high-standards presentation was actually like: presentations are about entertainment, not information; use well-designed and highly focused slides, not templates; only have one speaker present (in retrospect, duh).
The session made students immediately 10x their presentation game, with TKS teams making their way to top finishes in external and internal hackathons in the next month. The first objective of TKS Community Talks is to provide more opportunities for students to supercharge their presentation skills and train the ability to give high-standards talks.
Inspiration for the second and third objectives come from the fact that, what makes TKS special are the people in its community. As a member of the community, you’re surrounded by smart, curious, and ambitious people, striving to do impactful things and collecting insightful knowledge along the way. Countless students I’ve met are working on their own initiatives: apps, communities, organizations; and by virtue of the Innovate program, every student is working to become an expert in some high-impact emerging technology. Aside from this worldly knowledge, TKS is a also journey of mindset-shaping and personal growth — knowledge that, on an individual level, is even more meaningful and valuable.
Encouraging and empowering students to share this knowledge has a triple impact. Speakers solidify their knowledge of their topic and, in the form of the recording of their talk, gain a valuable portfolio piece. Community members get a high quality introduction to an interesting project, field of knowledge, or personal mindset. Finally, connections within the TKS community are built at scale where they wouldn’t have been otherwise: from speaker to audience, speaker to speaker, and audience to audience.
All in all, phenomenally. I’ll break things down by smaller events preceding the night of the talks.
I first announced TKS Community Talks on Sunday, November 22. “Talks open to the TKS Community, featuring…YOU!” went the premise. I included some incentives for giving a talk, and some general ideas for topics. I didn’t know what the reception would be.
The first big win was that this seemed to be something the community was interested in. More than 40 people reacted to the post saying they would be interested in giving a talk. Two and a half hours after sending the initial announcement, I posted a follow-up with a date for the talks (December 4) and a speaker application form. December 4 was a bad date for the event — the Innovate challenges were due on the same day — and after getting a few messages about it, I announced that Community Talks would be pushed back one week. The deadline to apply would be December 4, and the talks themselves December 11.
In the first two days, five applications were submitted. One week later, those were still the only five applications, and I began to get worried. I had seen countless initiatives requiring community involvement that had flopped because, no matter how many people said they would be interested, whether they would actually put in the effort to follow through was a whole different story. I messaged the Activate Slack seeking help or advice, but for the most part didn’t get anything valuable. I sent out a last reminder to the global channel.
On the Friday that applications were due, I had a conversation with Sabeeh that changed everything. I asked for advice about the talks, and this was the first thing Sabeeh told me: “community building is a lot of bitchwork.”
Translation: when organizing an event or organizing people, optimizations and shortcuts can never replace direct person-to-person work. With this in mind, Sabeeh and I compiled a spreadsheet of everyone who reacted to the initial announcement, and DM’d them asking if they were planning on applying. Immediately, a sense of control over the project returned. Five people said they were planning on applying, four of whom eventually became four of the seven speakers in the first round.
So how did hype-building an and speaker recruitment go? I got off to a strong start, but missed out on the key need to engage people directly. Thankfully, Sabeeh came through at the end to add this missing piece right in time, netting us nine total applications. It’s about a fifth of the people who initially expressed interest in speaking, but it was an effective enough funnel to make the first round successful. Interest and conversion rates across the funnel should only increase as the event becomes better-known, and as application deadlines don’t overlap with high-energy-requirement TKS challenges.
On Sunday, December 6, Sabeeh and I got to work. Community hype was there; nine applications were in; all that was left was, as Sabeeh would put it, bitchwork.
We came up with a timeline for speakers to prepare their talks: on Monday night, submit a high-quality title and blurb. By Wednesday night, submit a recording of the talk for feedback. Sabeeh and I considered the idea of a dress run, but Sabeeh pointed out that this would remove the pressure to have a talk actually well-prepared by then. Asking for a recording was a natural alternative for making sure talks were prepared.
At first, I was concerned about whether speakers would follow through with their deadlines. For the most part, though, speakers were really good about getting things done. DMs were quickly reacted or replied to, with titles and blurbs coming in on Monday and recordings on Wednesday (sometimes with an extra message to nudge).
I think the key reason that this was the case was because the speakers viewed this event as something “legit” that they had to be on the ball about. From my perspective, every aspect of Community Talks was created in the moment, and I had no idea if it would work. From the Innovates’ perspective, though, I was an accomplished Activate who knew what I was doing, and my messages were written in the same style and appealed to the same principles as messages from Nadeem or another director.
On top of this surface-level impression, I was passionate about making the talks successful and put in the work to push the speakers to as well. Even in my very first message, I pushed speakers to high standards, providing detail-oriented guidance and high standards examples. Half of this message to Neha, for example, is just about writing a good title and blurb:
This trend continued on Wednesday, when recordings came in. After chasing after the speakers to get their recordings in Monday and Wednesday afternoon, I suddenly found myself at 11 PM with a pile of them, and no prior plan for what to do. Sabeeh had allegedly recruited other Activates to help give feedback, but on Wednesday night at 11 PM Sabeeh was sleep and I didn’t know who had agreed to help.
I couldn’t wait, though. A 12-hour delay in feedback at this point would be losing 25% of the time between now and the talk. With the guiding idea in mind that “community building is just bitchwork,” I dove in, churning through talks and generating detailed feedback. I wasn’t overly intentional about giving in-depth feedback — I didn’t even have a clear idea of what feedback would be most helpful — but I had the passive recognition of high standards, and the passion driving me to to whatever it took to made this happen, to push through and make it work.
This is what those rounds of feedback looked like:
I sent two on Wednesday night, and three more Thursday morning. Sabeeh joined in on Thursday and gave the last two rounds required. If a speaker came back with a question or intermediate draft or recording, I replied as quickly as I could, bashing out in 10 or 20 minutes yet another round of notes.
For the most part, the potential of the talks was clear. The speakers had strong messages that they wanted to send, or really interesting technologies to talk about. In most cases, they knew far more than me about their topic.
Across the board, there were also issues in one way or another. Points weren’t ordered right; high-level concepts weren’t illustrated concretely enough; stories and examples were misaligned with the key message; slides were poorly made. The first draft of talks weren’t high standards. The speakers lacked the intuition that I, Sabeeh, and others have grasped after a year of intentional practice. But that was okay — the intention of these talks is to build those opportunities.
The second drafts, and to some extent even final talks, didn’t truly hit high standards in terms of overall presentation, either. But they got a hell of a lot better, and it was immensely rewarding to see my feedback be taken into consideration and thoughtfully implemented into their work.
The talks were well more than good enough to be watched at a community event, and the feedback was harsh and comprehensive enough to have pushed the speakers’ abilities. For a first pilot round, I consider this a success — it achieves all three originally planned objectives.
For future rounds, though, we can do better. I would say that the talks from this round aren’t good enough to be worth watching on YouTube. Two will essentially be portfolio pieces. The three technical introductions are great for community members, but don’t compete with the multitude of well-made videos out there on YouTube. Even Satvik’s unique, personal talk, which leads in views on the channel at more than 100, was still not polished enough to be watchable without qualification.
On Tuesday night (Wednesday morning) before the talk, I sent out the first announcement message for the talks themselves (i.e. not the signups). I wasn’t too worried about building hype, as there was plenty of interest before and sending out a simple announcement message isn’t hard in comparison to rallying a cadre of speakers, but it was something that had to be done.
The Tuesday announcement had three slides attached to it, featuring the titles and blurbs that speakers had refined and sent the previous day.
I was surprised by how well the slides came out (Navid even complimented me on them 😍). When I first set out to make them late, late on Tuesday night, I had no idea of how I would approach it. I wanted to hype the event and the speakers up, but I haven’t seen many conference posters before; I didn’t have a default layout to reference. Community Talks’ presence has been purely in the form of Slack messages and Notion pages so far, so there were no fonts, branding, or visual cues to reference, either.
On this last point, I realized that I was mistaken. I had used one of Notion’s default gradients in the original Notion page, so I copied it into Figma and started from there. I played around with fonts and added in the unicorn emoji until I came up with the all-caps, bold face “COMMUNITY TALKS” branding. Suddenly, things came together. From the chaos, the uncertainty of how I would make anything that looked remotely good, came graphics that looked…good.
I attribute this outcome to the high standards in graphic design I’d honed over the years, having done a ton of it often at high levels of scrappiness and beholden only to my own judgement. The impact of these graphics was that it lent yet another layer of legitimacy and high standards to the event overall, making the community more excited and speakers more motivated.
On Thursday, I made another graphic with the final order of the talks that Sabeeh sent out to #tks_global and other channels, along with the Zoom link. This was the last announcement, except for a final reminder Sabeeh sent shortly before the event on Friday.
Before I knew it, it was Friday. It felt strange given how much that day had been on my mind — now it was here. TKS Community Talks was actually happening.
Aside from last rounds of feedback, I didn’t start final preparations until 4 PM. The remaining todos were everything other than the talks themselves: the intro, the slides, the questions, the transitions. Sabeeh and I worked out last logistics — we would alternate on introducing speakers and handling their questions; our introductions would be well-rehearsed and high-quality; we would run on a strict 15 minute per speaker time table. We reached out to the speakers and told them to hop on the Zoom call early, at 6:30, to go over everything.
After Sabeeh left for another call at 4:30, I cobbled together a mini-talk for the introduction to the event — the vision and story slides at the beginning of this post. I knew what I wanted to say, but I wanted my own talk to be really high-standards and set the stage. I did what had worked for me during the Illumina challenge and wrote the script as I made the slides in Figma. At 5:30, I hopped into After Effects to export a quick 5 minute countdown graphic, then ran home to make the pre-show call at 6:30.
The script and the slides alike were high quality, in my opinion, but the timeline was simply too tight. I edited the script on the train ride home, not even trying to memorize it. I finished and linked up the slides while the 5 minute countdown played from 7 PM to 7:05 PM, the TKS community in the call watching.
At 7:05, I was theoretically prepared, but things just fell apart. I grabbed a second computer to pull up the script as I presented, but I didn’t have time to figure out how I was actually going to do it convincingly and ended up presenting purely from the slides, a recreation of what I wanted to say emerging from the imperfect recollection of script run-throughs. The transition to the Figma slides from the countdown was awkward; slides didn’t display properly, and two of them weren’t even linked, leading to me having to interrupt the presentation entirely to fix it.
In short, the intro to the event was a disaster. It was theoretically prepared at the high standards I held myself to, but the execution was completely botched.
Still, with stoicism and antifragility, or whatever in-the-moment unintentional versions of these mindsets I had, I pushed through and I made my way to introducing Zayn as the opening talk. The audience was excited, and I was more than happy to hand it off to Zayn’s much higher standards execution.
From there, it was smooth, smooth sailing. The speakers were well-prepared, with all technical issues worked out in that 6:30 call (unlike my own 😭). Their talks had improved even from the last recordings I had seen. There were maybe 30 people when the talks began. People kept trickling in, and the number increased to above 50 and stayed there for a good stretch of the night. The audience was engaged, hyping up speakers in the chat, and asked good questions. We made it to intermission right on time, then through to the end, thanking one more time the speakers and everyone who came to support them.
One idea Sabeeh brought up early on was to have a survey for attendees to fill out after the event, collecting valuable feedback for the future. In the last-minute rush, we had no such survey going into the event. But during the 10-minute intermission, I brought it up to Sabeeh again. It would be ideal to have the survey before the end of the event, so we could tell people to fill it out while they were still on the call (a trick I learned from Andover classes allotting time for feedback surveys during class rather than outside of it).
In the second half, Sabeeh got to work. During the last talk, I looked over the questions and cleaned up wording. By the time Neha’s talk was over, our survey was ready to go.
By the end of the event, around 30 people were still there. Thirteen of them filled out the survey that night, and five more filled it out from a Slack message on Sunday.
Q: How much did you enjoy this event? (1 = Not at all, 10 = Favorite TKS event. Average 8.83)
Q: How professional was this event? (1 = Not at all, 10 = TED. Average 8.67)
Q: How valuable was this event to you? (1 = Learned nothing, 10 = Life changing. Average 8.33)
In retrospect, we should have told respondents to be critical and hold us to high standards as organizers, too. I would assume that these ratings are somewhat inflated, but they do hold some good insights.
High enjoyment of the event was nearly universal.
More people were willing to criticize the event’s professionalism, which makes sense given the botched opening, technical difficulties, some connection lag, etc. Still, I was surprised to see that two thirds of all respondents gave our event a 9 or 10 on professionalism! Aside from our mishaps, there were definitely stretches of the event that we were able to execute on at high standards.
Personal value was the most evenly distributed of the bunch — still all above 7, but our talks aren’t delivering the educational or community-building opportunities that would make it invaluable on its own.
More specific insights come from open-ended questions about what went well and not.
Q: One thing that worked well? (required, answers condensed and aggregated. Some responses contributed to multiple items)
As I had thought, the speakers and their talks ruled the night. Respondents remarked positively on the topic selection, speakers’ enthusiasm, and the quality of talks.
Surprisingly, smooth execution on the part of Sabeeh and I was the second most-appreciated aspect of the event, with positive remarks on introductions, question facilitation, the slides we had running between talks, and an overall feeling of the event being well-planned and tightly executed. Aside from the slip-ups at the beginning, this also makes a lot of sense, and was something Sabeeh and I were definitely intentional about, even if we didn’t put a lot of time into it.
Publicity and hype-building were mentioned in three responses, and two speakers who filled out the survey appreciated the feedback that Sabeeh and I gave them.
Q: One thing that didn’t work well? (required, answers condensed and aggregated. Some responses contributed to multiple items)
A third of respondents had no flaws on their mind to jot down. One recurring answer related to technical difficulties — referring, I’m guessing, to the rough intro delivery — and connection issues, which interrupted Okezue’s talk periodically. There’s not too much that can be done about connection issues, and technical difficulties just require more practice to prevent from happening.
Two categories of answers related to audience interaction: many cameras were off, making speakers and audience embers alike less comfortable; and questions were sent by chat instead of directly asked by voice. The first point is one I was going to address by asking attendees to turn their cameras on, but forgot to talk about in the chaos of the intro. The second point I initially decided against because I wanted to keep our timing tight, but might be worth experimenting with in the future.
Confusion or dissatisfaction with the style of the talk was brought up twice. My opinion is that technical and more story-driven talks should both be encouraged, and can both be highly engaging to a general audience when done well. This then, again, is a matter of raising the quality of talks to the point where this is no longer a problem.
All other answers were mentioned by only single respondents. The event feeling like it was too long reflects that the talks weren’t engaging enough — you could easily binge TED talks for an hour and not feel bored, for example — and that perhaps our intermission should have been sooner. The optimal time for focusing is 45 minutes, after all, and our intermission was 80 minutes into the event. The talk order made it hard for those who could only make it for parts of the event to get a full sense of things, but I think there’s also merit to prioritizing the flow of talks for attendees who hear talks in sequence.
Q: Would you attend a second round of Community Talks in January?
This one speaks for itself.
Q: Would you apply to speak at a second round of Community Talks in January?
There’s plenty of interest in speaking in the second round, but again, I don’t expect conversion rate to be anywhere near 100% from expressing interest to sending in a high-quality application. A follow-up question about potential topics revealed that most who expressed interest didn’t have clear topics yet.
Q: Would you be willing to help organize future rounds of Community Talks, or help mentor speakers?
Willingness to help organize was less enthusiastic, but still significant. Putting together a trustworthy, high-standards team will require much more than just seeing who’s interested — more on this later — but it looks like at least a few quite legit people have expressed interest and will be worth reaching out to.
Certainly, there’s some bias in this survey. Those who think favorably of the talk would be more likely to put in the effort to complete the survey, for example. But there’s a big enough critical mass of positive sentiment that I feel confident labeling the first round of Community Talks as something that attendees generally thought highly of, specifically in terms of enjoyability, speaker performance, and organization. Targets for improvement primarily surround more preparation to eliminate avoidable technical issues, as well as even higher standards talks for better audience engagement.
On Saturday, I downloaded the 4 GB, 3000x2000 recording and painstakingly exported each of the seven talks from my laptop (each one took about 30 min). I appended a quick outro to the end, and prepended a few seconds of the intro slide for each talk to the beginning.
Later that night, I re-recorded the intro how I had planned for it to go, and also created some more eye-catching thumbnails. I changed around a few titles to be clearer and more attention-grabbing, too: “The Newest, Most Exotic Cuisine Is Grown In a Lab” became “How 3D printed meat can eliminate 25% of climate change,” and “Building a computer from cells to solve the biggest problems in medicine” became “How computers made of DNA will revolutionize cancer treatment and diagnosis.”
I’m not experienced with virality the way Sigil Wen and Michael Ye are — growth marketing, in fact, is one of the biggest weaknesses of mine — but creating some generally high-quality thumbnails and titles, that I was very competent at. The branded thumbnails I created are one last part of this first round I’m proud of:
I sent links to all the recordings in a last message to #tks_global and the Activate Slack. The channel now has 11 subscribers and the most popular video is Satvik’s at 140 views, so it hasn’t taken off in any way, but that wasn’t part of the vision I was pursuing. That’s not to say that it’s not worth pursuing; it’s a reflection that it wasn’t in my comfort zone, what I felt I could successfully drive for, but this is perhaps even more so a reason for me to pursue it. More on this later.
For the intentions of this first round, though, I don’t consider the lack of YouTube success a failure at all. The channel is now the permanent home for these talks, transforming them from ephemeral presentations into reliable portfolio pieces, if the speakers chose to use them so. They’ll serve also as examples for future rounds, potential marketing material, etc.
With all this review out of the way, what were the key successes, failures, and lessons learned?
Between the night-of survey and an interest form sent to various Slacks with recordings on Saturday night, 23 people have expressed interest in speaking for round 2. To be sure, out of 44 people who expressed interest in the first round, only 9 applications ultimately came in. Positive feedback is cheap, and doesn’t reflect who will actually put in the effort. But here I think the sentiment towards Community Talks is positive enough, and genuine enough, to demand a second round.
What would a second round look like? First, I think our objectives from the first round are still in good shape:
There are a few key points from the first round to keep up:
And now the fun part, what to improve.
There are two general outcomes on my mind:
This goal can be further broken down:
I expect the quality of speakers to continually increase, both as more legit speakers consider speaking in Community Talks, and TKS students overall get better at speaking. One actionable step to encourage higher quality talks is with a more comprehensive application process. This might include a 90 second pitch for the talk rather than just a text form, increasing the bar of entry and giving us a chance to assess applicants’ baseline effort. Interviewing applicants before accepting them is another idea.
Feedback is another aspect where we can do a lot more. In my opinion, giving feedback is actually one of the things Sabeeh and I did best last round, but only considering the scrappiness and tight timeline of the event. I probably didn’t spend more than 30 min - 1 hour total on any single applicant; all talks only went through one or two rounds of revision maximum before the final talk. There were plenty of improvements to be made on the final talks given. Feedback quality can be improved either by allowing more time for it (i.e. two weeks instead of one), or throwing more human resources at it (i.e. bringing in other Activates or even directors to help).
Regrettably, there was no single metric for “quality of talks” on the survey; the closest we got was professionalism of the overall event. Quality of talks is definitely a question to add in the future
Turnout has a clear metric in the form of simply how many people there are in the Zoom call. For the first round, that was around 50. For the second round, I hope it will hit 100. (There are hundreds of kids in Innovate…)
NYC Director Hayley once shared this tidbit. There are two sources of motivation: excitement (dopamine) and fear (adrenaline). Excitement comes from hyping up the talks and getting them as high-standards as possible. That much makes sense, but let’s also consider creating some fear as motivation 😈
It’s not as evil as it sounds. FOMO (fear of missing out) is a commonly desired effect to induce when building and marketing products. We definitely want to get to the point of building FOMO around Community Talks, where it’s so expected for everyone to go, or to give a talk, that TKS students feel like they’re missing something important if they don’t go.
We also have another fear on our side: fear of disappointing your directors by not making the most out of TKS. One thing missing from the first round of the talks was robust director support. Ian and Navid offered to help, and Steven gave a small shoutout a few hours before the event, but there was no formal announcement, buy-in, or anything else. For the second round, I want directors to help strongly push the talks: encourage students to apply, give feedback to speakers, and rally the community to attend on the night of the event.
My first thought after writing down these goals was that I need to execute on them NOW. The last sessions of the year will be this weekend (December 19/20); after that, the next session will be January 9/10. To stay on a monthly schedule, capitalize on the momentum from the first cycle, and leave two weeks for feedback, I imagined a Monday, December 28 deadline, and a Friday, January 8 talk date. Our big opportunity for director on-boarding would be in this weekend’s sessions, dropping in to give the pitch to apply to students directly, and asking directors to help more afterwards.
As I typed this, I messaged Sabeeh, and then Hayley. Sabeeh replied enthusiastically and similarly already started messaging people. Crazy how fast things can get moving, and how great it is to have a co-founder.
But then I thought more about it. If we were to go by this December 28 / January 8 timeline, the entirety of the talk cycle would happen during the holidays, sandwiched right between sessions, when kids and directors alike are most likely to be busy not with school or work, but with family and personal commitments. Talking with Sabeeh, I instead set a deadline of January 11, the Monday after the first session of the new year, with the talks themselves on January 22, the Friday of the following week.
We kept up the announce-this-weekend pace, though, setting up a spreadsheet and reaching out to every city director asking for five minutes at the beginning of each cohort session to pitch Community Talks. This is already be an ambitious initiative. Sabeeh and I can’t go to all the sessions, so we’d have to recruit other Activates and even Innovates to help cover all of them.
It’s exciting, though, to have this momentum going already, to get right back into checking off boxes in a spreadsheet. This momentum will build into the larger iterative changes we hope to make: directors being more involved; forming a core team of students to help organize; more and higher-quality submissions; more time for feedback; and ultimately, better talks and more people come to watch them.
TKS Community Talks was an idea that came to my mind in September, that stayed on my mind long enough to make it into a 1-on-1 with Nadeem two months later, and then — in the course of three weeks — suddenly became reality, involving dozens and dozens of students speaking or supporting. Needless to say, it’s been a really exciting initiative so far.
If we continue executing well, it’ll only get more exciting. I’m lucky to have Sabeeh’s support, incentivized by his community manager role. Zooming out, I’m incredibly lucky to be in a community with such a strong culture and with so many amazing people in the first place, that makes it possible to put on an event like this.
What keeps me so passionate about making Community Talks a success? As I wrote in a recent reflection post, “when I encounter a problem that I know I can solve, there’s no stopping me.” Community Talks seemed an obvious solution to build out, with a wide-spanning impact on the TKS community, yet nobody else was building it. I grabbed the opportunity when it first occurred to me, and there’s plenty of problem-solving left for me to do.
The long term goal is to make Community Talks a sustainable, high-value part of TKS; a foundational platform for training speaking skills, sharing knowledge, and building community.
The first-ever round was an exciting first step, offering compelling evidence that it’s entirely possible to make this long-term vision into reality. The second round will improve on the first, and so on and so on, on that unceasing march towards excellence…