You Won't Find the Perfect Co-Founder, and the Company Only Fires at Half Power
His judgment: giving up equity just to fill a co-founder slot doesn't pay, there is no perfect partner, and without one the company only reaches half of what it could; going solo forced him to learn everything himself, which turned out to be the biggest force multiplier.
The video won't play here. Listen to the audio instead:
The argument · tap a timestamp to hear it
Eight months of business plan, not one customer talked to
On January 28, 2020, at 14, he sat in front of a whiteboard at home and drew out Frentor — "Airbnb but for renting things," a tool rental marketplace — then spent eight months writing a business plan, which he says is absolutely not what he would do today. Back then research meant flipping through articles one by one, and with a narrow network and a fear of cold calls, he never talked to a customer from start to finish. The mistake he names isn't "making a plan," it's "shutting everyone out"; that method carried over to Omen, and the only thing taken off the block list was customers. Looking back, he says, a few more searches would have surfaced a whole graveyard of peer-to-peer rental markets, every strategy already tried by someone.
— Zach LabergeWhen he shut the company down, he took care of customers and team
Frentor ran for four and a half years, raised about $3M, had pivoted to software and GPS tracking for heavy equipment owners, had six-figure revenue, and was down to one or two months of runway. Canadian severance law is unusual — the company can't pay, so the individual is liable, and a 17-year-old could end up carrying hundreds of thousands in potential liability. He chose to sell the assets: customers handed over wholesale to another company, team members given several months of severance. To the same notification email, Silicon Valley people replied "waiting for your next thing," while the traditional Canadian circle asked "why didn't you keep going." His own calculus: time is the enemy — a company that ran four and a half years on a few million raised, investors will ask "give you another three million, what actually changes."
— Zach LabergeFor the second company, first ask where winning is easiest
After closing Frentor, the first thing he did wasn't look for a new direction, it was reconnect with the construction and rental company CEOs and executives he'd dealt with over the years — among them Adam Card, director of innovation at United Rentals. These companies had never bought Frentor's product, many had rejected him many times, but he'd always kept the line in reserve: when I have a next thing, can I come tell you about it. What he pitched Adam was a sensor mounted on the engine that predicts failure, and the answer was "if you actually build it, we'll buy it." A few weeks later he contacted a few investors and raised $3M in three days. His reasoning: going after rockets is a huge learning curve, he knew no one and wasn't a technical founder, and his strongest ability was sales; this industry he already understood to a dangerous degree.
— Zach LabergeCustomers only give feedback in the direction you bring them
Omen's earliest hardware was a vibration sensor mounted on the engine. But after running it with five customers, the feedback started converging on one point: engines don't really fail much, but when the hydraulic system fails it's 30% to 40% of the whole machine's value. Why did no one say so at first? Because the solution he brought them was the engine. His explanation: these industrial customers are great, willing to let you in the door, test, deploy, but most customers won't push you toward another problem on their own, because that isn't their business, they're just happy to support you. Asking further, "how do you handle it now," the answer was physically drawing oil samples, mailing them to a lab, waiting weeks for results. So the direction became real-time fluid testing on the machine — six months later a customer said "if you only build one thing, build the hydraulic sensor," and the next day they went all in on the pivot.
— Zach LabergeSix months early, reading the opportunity off Nvidia's temperatures
Looking outward from construction machinery, they screened several markets that also depend on fluid testing: chips, manufacturing (like factory monitoring for Bridgestone), and data center liquid cooling, which was just starting to heat up. The connection: the coolant used in data centers isn't that different from the fluid running through a CAT machine, and they were already monitoring it. He still used the guerrilla playbook, cold-emailing everyone, including the CEO of CoreWeave. Early feedback was bad — many said this isn't a problem, and a guy who'd spent 12 years in bitcoin mining laughed in his face. He estimates he was about six months early. What really convinced him was hearing that Nvidia planned to push operating temperatures to 34 to 43 degrees Celsius on its next-generation platform, and 34 to 38 degrees is the ideal environment for bacteria to multiply — like no one drinking from a pond that's been sitting. After that, large-scale liquid cooling failures started showing up.
— Zach LabergeA founding team is not the same as a co-founder
He separates "solo founder" from "having a founding team": a solo founder can still have very early founding team members who are genuinely all in. During Frentor there was a COO who stayed with him four years; early at Omen he recruited a CTO out of Nest's early founding team, and the two of them squeezed into his spare bedroom building hardware and mounting sensors on excavators. He agrees with Eugenia, founder of Replika and Wabi: founders reporting directly to the founding team is closer than putting a co-founder layer in between. He also names the downside — with co-founders there's always ego, because a CEO co-founder and the other co-founders are fundamentally different, and that tension sometimes becomes a hiring obstacle or internal friction.
— Zach LabergeAdvisors fail mostly because the cadence isn't fixed
With most advisors he meets a fixed 30 minutes every week, and the very busy ones move to every two weeks — the point is the cadence is locked in. His observation: any advisor who requires manually scheduling time each week eventually drops off, fine for the first few weeks, then a month in these people get busy. He also dislikes gathering advisors in one room to hear everyone praise him: some are embarrassed to speak up, some dominate, and it amounts to merging five people's one hour into one person's one hour, which doesn't pay. He uses a template Peter gave him, split into experts and a few middle tiers, spelling out meeting frequency, number of referrals, and the equity percentage by stage — for example, bringing on a very vertical expert at growth stage, 0.1% advisory shares, vesting over two years, plus a six-month flip. He basically doesn't pay cash, unless he only needs an expert for an hour or two, then it's $500 an hour.
— Zach LabergeBad outcomes are yours, and so are all the good ones
First the cost: it's lonely. Pressure from what investors, customers and you yourself promised all lands on you alone; the best founding teams will carry it with you, but the loan agreements and credit cards have your name on them. He also mentions a more hidden cost — whether the bank holds forty million, ten million or ten dollars, to him money is just a tool, but many people treat it as a pig to be divided up, which makes him prone to attracting contractors or the wrong people. The upside: bad things are yours, and good things are all yours too. Because he had no technical co-founder, he was forced to learn plasma physics and spectroscopy, things that otherwise wouldn't have been his to understand; and almost everything ultimately reports to him, so he has no choice but to get smart. He calls this the force multiplier.
— Zach LabergeIn their own words · checked verbatim
If you don't have the perfect co-founder, like that's when your business falls apart because you're only 50% as good as you could be.
Zach Laberge0:02
I probably put too many blinders on. It was just like, all right, I'm going to build this
Zach Laberge10:06
you really got to be critical of yourself a little bit and say like what do I have the highest chance to win at?
Zach Laberge26:15
most of these customers don't want to take you and like point you in a different direction because, you know, that's not really their business in some ways
Zach Laberge31:18
I've never heard of this problem. I've been in this space for like 12 years
Zach Laberge35:20
I think the world's moving towards more democratized liquid cooling versus away from it
Zach Laberge36:20
there's always ego cuz there is a fundamental difference between the CEO co-founder and everybody all the other co-founders
Zach Laberge45:26
almost every adviser I was like manually scheduling stuff with from every business like eventually falls off because these people get busy
Zach Laberge51:30
Figures
| Frentor total raised | about $3M | 19:13 |
| Frentor annual revenue at shutdown | six figures USD | 19:13 |
| Hydraulic system share of whole machine value | 30%–40% | 29:16 |
| Operating temperature of Nvidia's planned deployment platform | 34–43 degrees Celsius | 36:20 |
| Ideal temperature range for bacteria to multiply | 34–38 degrees Celsius | 36:20 |
| Omen's funding when it pivoted to data centers | about $13M | 37:20 |
| Advisory equity (growth-stage expert tier) | 0.1%, two-year vesting, six-month flip | 53:30 |
Glossary
- neocloud
- A cloud provider that specializes in renting out GPU compute; CoreWeave and Lambda in this piece are examples.
- preempted
- When an investor proactively issues a term sheet before the formal fundraising process finishes, grabbing the round first.
- advisory shares
- Equity given to advisors, usually priced by stage and vesting over time, with the amount varying by contribution tier.
- fluid testing
- Drawing hydraulic oil or lubricant samples from a machine and sending them to a lab to judge wear and failure.
How to listen
Founders who are going solo or about to, investors looking at early-stage deals, and engineers who want to know what AI data center liquid cooling is actually solving.
At minute 10 there's a recruitment ad for this show itself; the childhood business history in the first 18 minutes can be fast-forwarded.