camps@betaworks

botcamp, betaworks first accelerator program, ran this summer. The program was a success and we’d like to share some data points and learnings gleaned.


When we announced botcamp in April, betaworks had never run an accelerator program. Though we had been immersed in all things bot for a while, Matt and Peter only first discussed the idea at Launch Festival in early March. From there, we embarked on a whirlwind ad hoc journey, culminating in the botcamp summit and demo day last month.

From the initial announcement, to the application (via bot, naturally), to the screening process, to welcoming and working with 8 amazing teams at the betaworks studio, we were learning as we went, albeit alongside some remarkable partners, mentors, and founders.

This was all, of course, magnified by the extremely swift and volatile hype cycle surrounding all things bot during the spring and summer. (Side bar, and perhaps the subject of another post: 2016 has been a wild, wild year).

When we announced the program, bots were the new apps and the next frontier. But, by the time the 8 participating companies arrived, bots were over-hyped, useless, and definitely not the new apps (even among their initial disciples).

Yet, despite that relative inexperience and healthy bit of clamor from critics, botcamp proved to be a resounding success. We could not be more happy with the companies, with the program itself, or more interested and bullish about how this emerging new interface is developing. As John and Chris discussed at the botcamp summit / demo day, we are in the formative stage of what is a new way of interacting with computers.

On the company screening side, we received over 450 unique inquiries via our bot, with 346 fully-completed applications. This highly competitive applicant pool meant that we were able to select a truly amazing group of teams (3% acceptance rate) building diverse businesses from a wide range of backgrounds:
- 4 serving businesses; 4 serving consumers
- 4 nationalities represented
- 3 from CA Bay Area; 3 from NYC; 1 from Toronto; 1 from Boston
- 3 teams female-led; 2 led by underrepresented minority groups
- Deep and wide-ranging professional experience from AppNexus, Bauman Moscow State Tech, The French Ministry for the Economy, Google, MIT Media Lab, Twilio, and XOlator, among others, and including numerous entrepreneurial ventures (with exits).
- Products based on years of academic research or on a weekend hack.

The teams with Phil Libin of General Catalyst

The 8 botcamp companies are: 
Bites — messenger-native platform for interactive daily digests.
CareerLark — performance feedback and employee engagement apps for enterprise messaging platforms.
Coach Otto — AI for human resource training through conversation.
Hugging Face — your personal AI friend.
Olabot — platform that makes it fun and easy to create a personal bot.
Statsbot — bot that provides a conversational interface into analytics and business intelligence.
Warren— a sidekick for creative freelancers to help manage their business.
Zoom.ai — the operating system for your workday through an automated virtual assistant.

But all that was just to get the teams in the door. Starting July 11th, we had to actually offer the founders valuable experiences and connections to help them build their companies…

Uh oh! Will they show up??

We did this with our function-specific, “platform” teams here at betaworks (investment, finance, product, data science, design) but also, perhaps more importantly, with the teams building bot-related companies here at Poncho and Dexter.

We also did this through an incredibly talented group of partners and mentors, ranging from deep technology providers, to nearly all of the major messaging platforms, to various media outlets, to a myriad of sharp, focused, and fiercely thought-provoking investors.

Finally, we should mention that we did this through investing, along with Jesse and Mike from The Chernin Group, $200k into each of the participating companies. As someone recently told me, “that’s a bit better than spending the summer zip lining and making lanyards.”

All told, we certainly threw a TON at the participating companies. From the beginning, we viewed these founders as our customers and, as such, one key measure of our success is how they experienced the program. Well, here’s what they said:

1 = “inspiring and encouraging” 5 = “intimidating and distracting”
Betaworks felt like it really was a product-first company and that’s really hard to find in NYC. I also liked the velocity of the events throughout the week, as much as we probably all complained during the process. It was awesome. I wish there were another 10 weeks. — botcamp founder
1 = “made perfect sense” 5 = “were a waste of my time”
I would replace the platform week with a group session where the teams talk through their designs and tech stacks. I think this would eliminate some of the weakest content while also facilitating deeper & more valuable interactions across teams. — botcamp founder

The above responses are selected from a much larger set of questions we asked the companies at the end of botcamp. While very positive on the whole, they do highlight a few areas where we did not do enough:
1. Making sure every session was well-prepared and tailored to specific, functional needs of the companies.
2. Providing more structured time for the teams to interact with each other to share knowledge, best practices, and mistakes.

By the time Demo Day arrived, it really felt as though the botcamp teams were experts in this nascent and fascinating space (just look at the way they held their own on stage with some tech and media giants). While we could have done more to facilitate knowledge sharing amongst the group while they were here, rest assured: the botcamp bond remains strong. We’re continuing to work with all of the companies to build their businesses, individually and as a group. To echo, here’s one more founder review:

Overall it was really a great experience, and we look at it as the beginning of an ongoing relationship, rather than as something that is done. It was hard to fully take advantage of the opportunities presented to us within the time limit of botcamp, while trying to build stuff, but hopefully seeds planted during botcamp can be harvested in months and years to come. Overall — thanks guys!

And, on that note, we certainly didn’t go it alone:

Ok, so given the success / first-timer luck, where do we go from here?

Well, there’s definitely room for ongoing botcamp-type programs here at betaworks. That said, we’re not about to pivot into becoming an industrial scale accelerator, either.

Rather camps, or thematically-driven micro-accelerator programs at the betaworks studio, should provide an efficient way for betaworks to invest in and work with promising early-stage companies while taking advantage of our existing studio infrastructure.

At their core, these camps should reflect and complement the evolving themes around which betaworks’ is both building and investing. The initial TechCrunch announcement really nailed this goal:

This mini-accelerator, with a full application process, is a different path to a similar result. At the end of the day, Betaworks is interested in clusters of companies that can help each other, and botcamp is one facet of that.
 — Jordan Crook

And one more thing: the botcamp program itself provided significant opportunity for betaworks to strengthen our network by inviting both new and existing partners, co-investors, and portfolio companies to participate. From the beginning, betaworks was designed to be an innovation hub for startup activity and company creation here in New York. We think that camps are a great part of that ongoing mission.