Andrew Berman

The through-line

Everything Andrew Berman says about company building comes back to one impulse: he wants to bring into existence the product he can see but cannot buy. He describes sitting on the investing side of the table at Norris Venture Partners, backing consumer and enterprise companies including video businesses like life-size and blue jeans, and concluding that he wanted what the founders had. "I was sitting across the table from these incredible entrepreneurs who were putting it all on the line and starting these great companies I wanted to do it I didn't know how I didn't know what I wanted to start" [2]. That gap between knowing you want to build and knowing what to build gets closed the same way twice in his account, by living inside a broken experience long enough to be certain it is broken. With Vowel the formulation is explicit: "the product that I want to exist has yet to be built and I'm building it today" [2].

The second constant is that he treats conviction as something you pay for in rejection and in time. He counts the noes, he counts the months of runway, and he counts the development cycle, and he is unromantic about all three. His summary of most founding stories is that "it's a zigzag line to that IPO" [2].

On leaving investing to build

His career reads as a deliberate walk away from the safer seat. He started in Silicon Valley in 2007 at a firm that "went bankrupt in 2008," then joined Norris Venture Partners at the end of 2009 doing consumer and enterprise investing, which he calls "an incredible experience learning" [2]. The venture job did not satisfy the urge, it created it, and business school was the bridge: "when I left to go to business school that was my plan I was going to start a company I wasn't sure if it was going to be two years after 10 years after but it was something that I always knew I wanted to do" [2]. He is candid that the investing background gave him craft rather than immunity. He knew how to build a deck, how to run a priced round, how the venture process worked, and "how to create a narrative," and still: "until you do it yourself for the first time and until you get those rejections you really don't understand what it's like to be in someone's shoes" [2].

The same restlessness explains his exit from Nanit in 2019, and he frames it as leaving from strength rather than escaping trouble. The company was on good footing, the management team was built, and "I just had a burning desire to start something new" combined with a market he thought would be a real challenge [2].

On the future of meetings and video

The Vowel thesis starts from a personal failure of tooling. With a team in Israel, a team in New York and people scattered across the United States, "collaboration was so hard uh and we tried all the tools Under the Sun" [2]. His diagnosis of video conferencing is that the underlying media quality solved itself while the product stood still: the audio got good, the video got good, but "the product experience hadn't changed since 2009" [2]. Vowel's answer is to "turn meetings into searchable shareable clippable audio," so that when a team in New York meets and the engineers are in New Zealand, "the second that meeting's over they'll have an instant artifact that's available to them" and can get the tldr of what mattered [2]. He describes customers giving the product their own name for this: "they call it time travel" [2].

Underneath the feature set is a claim about what video now is. He argues the category has been misfiled as a communication medium when in fact "it's how we work," whether a team is remote, hybrid or in between, and that the platform shift of a few years ago left people on video for hours a day in an office and potentially all day when distributed [2]. Hence his framing of the goal as "enabling users to collaborate not communicate," with the product suite integrated into other tools and systems of record, and his ambition stated plainly as building "the next great collaboration company" [2].

On hardware being hard

He is one of the few founders who can hold hardware and software up against each other from direct experience, and his comparison is about the speed of the feedback loop. At Vowel a change takes a week or two, the team ships weekly, and a coffee with his head of product in the morning can turn into something live next week [2]. At Nanit the app could ship every few weeks, "but the app's only one piece of it," and the actual hardware development cycle runs "between like 18 and 24 months," so a change to the core camera, the mounting or the industrial design lands two years out [2]. The consequence he cares about is emotional as much as operational: in software "the instant gratification the um the user feedback comes a lot faster" [2].

The second penalty is capital. "There are only a few investors out there that will fund" hardware, and while "everyone says they do," when you "dig down deep it's few and far between," whereas essentially every well-known venture firm is comfortable with software [2]. He also has an unusually concrete picture of what happens after the design is done. He has walked factory floors in Guangdong outside Shenzhen and around Shanghai, including one plant making HP printers, Apple iPhone chargers, specialized parts for Nike and the Xbox, with LCD monitors overhead showing the utilization of every line [2]. His organising distinction there is that "there's high volume and then there's High touch," and that building six million-dollar machines a month is a fundamentally different discipline from rolling ten thousand iPhones off a line in a day [2].

On fundraising, grit and the number of noes

His fundraising advice is arithmetic rather than inspiration. Before Nanit's institutional seed, the company raised roughly half a million to three quarters of a million from friends and family, and he is precise about what that felt like: burning about $25,000 a month and having to "go raise fifty thousand a month more or less" to keep the lights on, which in practice meant "go take 30 meetings and hope you have a 10 conversion rate," repeated for six to eight months [2]. He also adds up the rejections rather than forgetting them, putting the total at around 350 separate investors across VCs, angels and super angels in the first two and a half years, and turns that into the line he gives other founders: "it takes one yes and until you've really gotten 300 no's a you're not really out of the ball game" [2]. Grit, in his account, "is what's what allows you to survive," and he points at Elon Musk as the extreme case while noting that a few founders find lightning in a bottle and most do not [2].

He is equally clear-eyed about the composition of a cap table. Nanit's pre-seed came from his business school friends, a group that includes people senior at a large cap buyout firm, someone running a venture and growth equity fund, and a public markets growth investor, so "people really understand the domain" and everyone shares deal flow [2]. He resists reading that as bragging rights, noting the class is full of successful people and that what his friends really did was watch "this uh this idea come from a pitch deck to a very large brand sold throughout the United States" [2].

On product prioritization and user research

He calls prioritization the hardest thing early on: "one of the most difficult things in the early stage that the company is prioritization" [2]. His working test for whether a signal is real is a low bar deliberately set: "if you ask 10 people and you get six to seven answers that are remotely remotely similar that's what a product manager would tell you is enough direction to know that you're prioritizing things correctly," and surveying six to ten people is a good enough sample if the commonalities show up [2]. He does not claim the craft as his own, describing himself as "not a classically trained product manager" and crediting his head of product at Vowel, Anna Marie Clifton, alongside another product manager, with continually running user research and hunting for trends [2].

At Nanit the research discipline was built in from day one with a user researcher on staff, plus heavy user interviews and prototyping, but the strategic bet was different in kind. Baby monitors were an established category, so the question was never demand, it was "how do you provide a 10x better experience," and the early hero message on the homepage was "the Tesla of baby monitors" [2]. Only then does the feature argument begin: sleep tracking, breathing in the app, milestone development, and which of those ships first [2].

On teams, technical risk and the stages a company passes through

The founding of Nanit turned on a specific reading of technical risk. His co-founders came out of Applied Materials, where they had built a computer vision electronic microscope for wafer inspection, million dollar machines that spotted semiconductor defects better than a human inspector, one having written the algorithms and the other the back end system, and one was doing a postdoc at Cornell Tech in a program explicitly designed to spin out companies, "sort of creating a mini YC program" [2]. From that he concluded, "between the two of them I knew we could build production grade Electronics" [2]. What he did not know was whether they could build a great consumer product experience, and he is direct that this was the trade he wanted: "was that a risk I was going to take every day of the week" [2]. When they announced in 2014 that they would use computer vision to track a baby's movement, "people thought we were crazy" [2].

He rejects any tidy story about early roles. "Everything evolves over those first few years in terms of roles and responsibilities anyone who tells you anything different I mean I think is lying," and at the start his own job was fundraising, product design, branding and recruiting at once, "you name it" [2]. Specialisation arrives with scale, until the company "starts to look like a normal company on the inside" [2]. He describes the arc not as a single milestone but as a sequence: product development, shipping, refining, shipping again, scaling the business, then continual product development, with the classic pivot point being the move from searching for product market fit to throwing gas on the fire in sales and marketing [2]. For a direct to consumer business like Nanit, he notes, that mostly means marketing dollars, because outside Amazon and a few big box retailers "there is only a few channels left to sell through" [2]. He credits Nanit's current CEO, Sarah, as an amazing force in building and scaling the company on the marketing and sales side [2].

Takeaways

• Judge video conferencing by the product layer, not the stream: audio and video quality improved while "the product experience hadn't changed since 2009," which is where the opportunity sits [2].
• The point of recording meetings is the artifact, so a distributed team gets "an instant artifact that's available to them" the moment the call ends, which customers described as time travel [2].
• Hardware and software differ mainly in loop speed and capital access: an 18 to 24 month hardware cycle versus weekly shipping, and far fewer investors who genuinely fund hardware [2].
• Count your noes before you quit: roughly 350 investor rejections in Nanit's first two and a half years, and "it takes one yes" [2].
• Early fundraising is a treadmill with a conversion rate, in his case 30 meetings a month for a 10 percent hit rate to cover a $25,000 monthly burn [2].
• Enough signal to prioritize is modest: survey six to ten users and look for six or seven roughly similar answers [2].
• In an established category like baby monitors, demand is not the question, a 10x better experience is [2].
• Expect roles to churn: "everything evolves over those first few years in terms of roles and responsibilities anyone who tells you anything different I mean I think is lying" [2].