Venture capital investor offers tips for IoT startups, says voice, AI are the fundable technologies of 2018
SANTA CLARA, Calif.–Founded was in 1911 by Carnegie Steel founders, Bessemer Venture Partners is the oldest venture capital firm in the U.S. The firm has invested in everything from PCs in the 1980s, to social media in early 2000s.
Vinay Iyengar, venture capital Investor from Bessemer Venture Partners, gave an audience of IoT developers at IoT DevCon in Santa Clara, Calif., tips for finding investors. Bessemer invests in seed funding through A and B all the way to the growth stage. The company invests in technology in consumer, health and the enterprise space.
Machine learning and artificial intelligence is a big draw for many investors now. “People around the world are building remarkable products around AI,” Iyengar said.
He added that now is a good time to get funding. “The capital infrastructure is there,” Iyengar said. 2017 was a really major year when AI emerged from the lab setting and is appearing in commercial products. “We saw real commercial applicability of AI and we saw a lot of AI products start to get funded,” he added. “Private funding almost quadrupled in 2017 for AI startups.”
Of particular interest is voice — as in, voice-controlled, screenless platforms, such as Alexa or Siri, where you use voice commands instead of text to search or do other tasks. Last year saw the arrival of screenless technology, with 25 million Alexa devices sold by end of 2017, and 25,000 are number of skills created by third party developers for Alexa. Iyengar says these numbers closely parallel what happened with the iPhone in 2010.
“I see a real platform shift here, a platform shift towards voice. The Alexa is to 2018 what the iPhone was to 2010,” said Iyengar.
Every year Bessemer produces “laws” — like a roadmap — for a specific technology. These laws are supposed to help the investors find the next big unicorn, a startup with a valuable over a $1 billion. Although Iyengar readily admits that some of Bessemer’s best investments broke their own laws, these rules can help developers understand if their company is fundable.
One way to be fundable is put “.ai” into your domain name. He was only half joking about this.
Here are Bessemer’s six laws for machine learning. These are what Bessemer looks for in a company they consider fundable.
- Have a data moat: your company needs to have its own proprietary dataset that has unique insights to the end user. Moats should be growing over time.
- Leverage existing tools: Many tools and APIs are out there; don’t reinvent the wheel.
- Do one thing and do it well.
- Enablement versus replacement: Don’t try to replace humans (yet). Try to improve something they already do: super-charge the human, but don’t replace their job yet. The reason: it’s too hard a technical problem to replace a human when you don’t even know if your product will sell yet. (Focus on the vertical issue and avoid the giants, said Iyenar.)
- Voice is the new user interface.
- Expertise matters: Get experts to work on your project. (The race for AI talent is fierce now, with PhDs in AI or related experience making $300-$500K a year.)
