Drones Rise as Chatbots fall off: YC trends for 2016
Every 6 months or so I get to come back to Mountain View and see somewhere around 100 startups present. My partner at Ringleader Ventures, Mike Bechtel attended the “real” Demo Day this year, which is a little more formal while I get to attend the practice — Alumni Demo Day. The crowd is a little more raucous, the founders seem a little more relaxed and there are huge cheers anytime a founder says the magic words: “And we are profitable.”
At either event, its always fascinating to see what trends are represented. Y-Combinator is probably the best high pass filter for seed stage startups, ensuring greater chances of success (or at least survival) versus what I see in the wild. Because of this, I am always fascinated to see what these “allstar rookies” see as the future. While there are mainstays from batch to batch, the rise and fall of topics and industries is always interesting. Here’s what I saw last week:
Deep Learning / AI
I’ve now seen 8 demo days, and about 800 startups of the 1181 that YC has backed. Machine Learning has always been a theme of the batches, but it has continued to grow and the specific method du jour changes. In this batch, two promising flavors seem to be rising to the top: Deep Learning and Computer Vision. Deep Learning (https://en.wikipedia.org/wiki/Deep_learning) is both a buzzword but also a specific methodology for using multiple layers of processing to extract valuable insights from vast data pools. It seemed to be the synonym for Machine Learning or Neural Networks in this year’s pitches.
Computer Vision is actually frequently paired with Deep Learning. It enables both fun things like the app Prisma (not YC but awesome) as well as counting cells in your blood stream, containers in a shipyard or potatoes in a field. These are all interesting applications because they are at the edge of what we thought was only commercial viable for huamns to perform in the last year or so.
It seems like Drones, at least as far as startups are concerned have moved into the slope of enlightment phase of the Hype Cycle (https://en.wikipedia.org/wiki/Hype_cycle). A few years ago there was a string of 2 or 3 consecutive demo days where Drones were everywhere. Some of those startups are still around, but a number also ended up being either too early to the space or bringing novel technology to a space where it was never going to beat out the incumbents.
Now that drones have infiltrated commercial and enterprise applications, this year’s batch of startups was the most interesting to me both personally and through the lens of our RLV investment theory. There are only a handful of founders and teams focused on the space, but they were focused on regulatory issues (collision avoidance), known problems for using drones (drone shields) and specific applications (security bot). Each of these seems to be addressing a great problem, the only question is who will win in each sector?
Gross Merchandise Volume (GMV) is growing as the buzzy metric for re-sellers and marketplaces. It’s an interesting metric because it seemingly combines both user growth and revenue into one number. While it might be the top-line metric for marketplaces, my midwestern sensibilities yearn to just see the revenue as a raw number. Especially right now, revenue and survivability matter to every startup more than anything else.
Chatbots go “Poof”
At the last event in March, there were either 6 or 7 chatbot startups. Here at RLV we think chatbot platforms, tools and implementations are a huge deal. Their applicability in voice control, phone replacements and limited bandwidth markets means their potential footprint is huge. While last time felt overwhelming, I am hoping we see some return in the future much like we have seen with drones, where the startups are presenting not just possibility but real application and nuance.