KubeCon Recap, Part 1: The View from the Vendor Floor

So, in case you missed it, KubeCon happened. Between holidays and my intermittent posting schedule, I'm probably going to spend the rest of the year covering the event. Since next week is Thanksgiving in the US, I definitely won't be posting then, so I wanted to get at least one post in before some of the buzz from the conference dies down.

Unlike last year, where I spent most of my post-conference blogging covering the various talks and such that I liked from the conference, this year I'm going to start off with some more business-y aspects of the conference. That's because, unlike any previous conference I've been to possibly ever¹, I spent 80% of my time on the vendor floor. It was definitely overwhelming, but I had a lot of good conversations and leads to follow up on, which feels exciting. So I'm going to start my blog coverage talking about the vendors.

The prep phase: getting stickers printed

As you may or may not have noticed, I've been in a big marketing and sales push recently, both on the blog and outside. I knew that I was going to be spending a lot of time talking to vendors, but I also knew that a lot of other people were also going to be talking to the vendors, so I wanted to make sure that I had the appropriate marketing to "stand out from the crowd". It helps that (as best as I can tell), I'm one of the only people in the industry trying to do Kubernetes simulation. I think it also helps that SimKube is a pretty memorable name—short, sweet, and to-the-point. So I just wanted to capitalize on those factors.

I've been following along with Michael Drogalis and learning from some of his experiences, and one of the things he mentioned is getting shirts printed with a QR code on the back that led to his product. I thought this was a great idea, and since SimKube now has a very cute and memorable mascot², I figured putting that on a t-shirt and wearing it around the conference could lead to some good conversations. So I printed off a few t-shirts for me and some friends, and we all wore them around. The one change I made from Michael's experience was, instead of putting a QR code on the back, I included a conversation starter: my goal was, I really wanted to get people to talk to me rather than just going to a website.

Unfortunately, as best as I can tell, the t-shirts did approximately nothing for me. I never had anyone comment on them or ask me any questions based on the prompt, and the friends that I had wearing shirts also reported similar results. So I was a bit disappointed by that. But I don't consider it a complete loss: I still have the t-shirts and I can use them for similar events in the future, where maybe they'll have different results.

The other thing that occurred to me wayyyyy too late in the process is that I should have stickers and business cards that I can hand out to people. Fortunately, I was able to find a place in town who could get me stuff printed in time for the conference, so I threw together some quick designs in Photoshop and got them made.

I made two big mistakes here, though: first is that the stickers didn't include the name of my product anywhere (I made stickers of Mr. Squidler and of the SK logo, but neither one said SimKube). So once I got to the conference I didn't feel like I could just hand them out or leave them lying around, because nobody would know what they were. So, fine, I handed a few out to my friends and kept the others for later³. The other mistake was on the business cards, I included a link to the SimKube website, and a custom email address people could use to contact me, but I didn't actually put my name on the card. This was sortof deliberate, because I wanted people to focus on the product rather than who I was, but after the third or fourth person asked me why my name wasn't on the cards, I began to realize this was a mistake.

All in all, though, for a few hundred dollars of marketing budget, I feel like it worked out OK. It brought at least some value, and I learned some stuff for next time. And then, of course, I spent the evening before I was supposed to fly out rather sick, and I almost didn't make it to the conference at all.

Who sells to the salesmen?

The first two days in Salt Lake were my "talks and networking" time; I gave a talk at Cloud Native Rejekts and met some cool folks there. I also spent time volunteering at the Contributor Summit. I'll talk about those events in a future blog post; today is all about the vendor floor!

My goal going into the conference was to talk to every vendor who was there. I quickly realized that a) this was unrealistic of me, and b) a waste of my time. There's a ton of vendors doing, idk, eBPF things or security things or CI/CD things, who just wouldn't even know what to do with SimKube. That doesn't mean I couldn't have talked to them, of course, but there were also just soooo many vendors overall that I needed to prioritize. After the first day, I had more-or-less talked to a third of the companies that I thought looked interesting, and that was a small fraction of the total vendors on the floor!

I spent a bit of time the first day perfecting my script: I would walk up to a vendor, ask a bunch of general questions about their product, and then figure out (in my head) where I thought SimKube would be most useful for them. Then I would ask a couple of really specific targeted questions, like "How do you solve X problem right now?" or "Have you ever considered doing Y?" I tried to tailor the questions to be a bit more technical, so that the sales rep would have to pull in a more technical person (at the smaller companies, this often ended up being the founder or CTO). I could tell that they didn't quite understand where the conversation was going, because these weren't the "normal" questions that people asked. And then once I had gotten them talking about the problems they'd identified in their own platform, I'd whip out "Have you ever considered trying to solve this problem with simulation?" and I'd hand them a business card⁴. This conversation model worked around 90% of the time⁵, and it got a lot of folks really excited. So I felt good about that—and, despite not including my name, the SimKube business cards that I printed ended up being the best marketing investment I made. I printed 50 and gave almost all of them away. I also made sure that I took pictures of badges of everyone I had conversations with, and connected with all of them on LinkedIn after the fact⁶.

By the end of the conference, I think I had talked to around 95% of the people that I identified as "priority 1", and my goodness it was exhausting. But I ended up with a strong set of really solid leads that I'm now attempting to follow up on.

So did it work? Who knows! I'll tell you when I land a client. But I'm cautiously optimistic that at least one of these leads is going to turn into something.

But what about the talks?

It was a bit sad to not get to go see more talks, because those are my favorite parts of the conference experience. Aside from my own, I went to two talks at KubeCon proper:

• Choose Your Own Adventure: The Observability Odyssey by Whitney Lee and Viktor Farcic: this was a live continuation of a long-running choose-your-own-adventure podcast series about Hero, the poor pod who just wants to get to production. It was truly a delightful presentation, and I really recommend watching the whole series. This talk was unique in that the two co-presenters more-or-less didn't coordinate ahead of time. They didn't know what the other person was going to be doing or saying at any point, and—this was definitely a bold choice, and I think maybe only possible by these two people! Whitney and Viktor have such a delightful chemistry on-stage, and it was a real joy to get to watch them as they deployed a bunch of popular observability tools for Hero.

• WASM + KWOK Wizardry: Writing and Testing Scheduler Plugins at Scale by Dejan Pejchev and Jonathan Giannuzzi from G-Research: this talk was an obvious one to go to for me, because they're talking about scheduling and using KWOK, which of course SimKube uses internally as well. It was also a pretty good talk, although a lot more niche. I did learn from this talk that you can now write kube-scheduler plugins using WASM, which is a pretty exciting development! No more re-compiling the scheduler from source just to get your own special plugin behaviour implemented. They also demoed the kube-scheduler-simulator, which is specifically designed to help you understand what the scheduler is doing and why. It's a little different from SimKube (in fact, I would argue that it's not really a simulator at all) in that it's just providing some additional annotations and observability into the scheduler's actions, whereas SimKube has a more expansive (but also harder to reason about) purview.

My own two talks (which I linked to last week) went well, too, and I got a lot of good feedback and engagement from them as well. So all in all, I'd say it was a successful conference! And, a week later, I'm maybe 95% recovered and willing to leave my house again :joy:

I hope those of you celebrating Thanksgiving next week all get some rest and relaxation time, and I'll be back here in a couple weeks with more KubeCon coverage!

Thanks for reading,

~drmorr

1. The two exceptions being SuperComputing 2008 and the American Chemical Society conference in Fall 2008, both of which I worked a vendor booth at, and were frankly miserable experiences.

2. Mr. Squidler says hi!

3. If you want any stickers, let me know!

4. The moment when the vendors suddenly realized they were being sold to instead of doing the selling never failed to fill me with an incredible amount of glee and satisfaction.

5. A few times I identified partway through the conversation that they wouldn't actually be that interested in SimKube, and at that point just pretended that I was a mildly-interested passerby, disengage, and move on.

6. I also let everyone who asked scan my badge, which means I'm now dealing with the deluge of spammy marketing emails from all the other vendors at KubeCon.