Devon Galloway on What Venture Money Is Actually For
A CTO turned investor on fund math, formidable founders, and the long road after the validation pyramid
TLDR; Devon Galloway co-founded Vidyard out of Waterloo, took it through Y Combinator in 2011, and has spent the fifteen years since both building it and, from 2013 on, investing through Garage Capital. He joined us to talk about the part most rooms skip: what happens after the validation pyramid, and what venture money is and isn’t for. The throughline was discipline about reality. The best pitches are dead simple, most software no longer has a durable moat, half of any good fund’s bets fail, and the thing that reliably wins is a founder who will figure it out. It paired cleanly with the workshop on default dead or alive, because Devon’s spent a long time deciding which companies are worth betting on, and for how long.
The first three dinners were founders talking about building. The fourth brought the conversation to the other side of the table. Devon Galloway has built, he is still building as CTO of Vidyard, but he has also written close to 300 early-stage cheques through Garage Capital, and that double vantage point shaped the whole evening. He started angel investing in 2013, fresh out of YC, back when he and his co-founder were paying themselves two thousand dollars a month and felt rich.
Fifteen years past the finish line
Devon’s least favourite founder trope is the one that treats getting into YC as the finish line. He called it the validation pyramid: the talks, the startup-school advice, the scramble to put a logo on a slide. For him and his co-founder, YC was the beginning of the beginning. The fifteen years after the batch taught him far more than the batch itself did, and would have changed plenty of his early decisions if he had known what was coming. The road is long, he wanted the room to understand, and the validation you chase early is rarely the validation that ends up mattering.
What those years did teach him is how to recognize a good pitch, and his test is almost aggressively simple. He and his co-founder day-traded together as co-op students, badly, and the tips that stuck were always the clean ones: a friend leaning over to say, here is the one thing about this company. He still remembers a tip about a small pharma company whose FDA clearance had been pulled and looked likely to come back. One sentence of insight, and the stock ran.
A startup pitch should sound the same. There is usually one real insight, and everything else is detail. The Groq pitch was that clean: AI is going to be enormous, the chips built for it are bad, I will build a great one, and I happen to be very good at chips. He gets three- and four-page founder emails all the time, and the length usually tells him the founder doesn’t yet know what is special about the business and is hoping the reader will find it for them.
Default dead, on purpose
Groq is also the cleanest illustration of the workshop’s theme. It was default dead for the better part of a decade: little real revenue, selling mostly to universities and research labs, burning capital to build genuinely hard technology. Nvidia bought it last Christmas for twenty billion dollars, the best outcome in the fund’s history. That is the exception that proves the rule rather than a counterargument to it. Most deep tech is default dead for years, and that is acceptable only when the prize at the end is enormous. Default alive, the frame from the earlier workshop, is simple arithmetic: on flat expenses, with the cash you already have, does a dollar in come back as at least a dollar? Knowing that number is what gives you a direction to climb toward, and most founders avoid running it because the answer is uncomfortable.
For companies that are not building a chip, the question of when to raise has a cleaner answer. Devon’s tell for product-market fit in software is the ratio of customer lifetime value to acquisition cost: put plainly, your company is a machine that turns one dollar into five, and if that is true and it scales, putting more dollars in is close to a no-brainer. The harder truth is that technology moats have thinned. Someone can prototype your software over a weekend and your margin goes to zero. The advantages that compound now are customer ones: knowing a problem better than anyone else, becoming the system of record, and getting genuinely creative about reaching customers when every inbox is buried in noise.
He quoted Buffett, that you find out who is swimming naked when the tide goes out, and the tide is out on software right now. Then Larry Smith: if you can crack go-to-market while it is hard, you should be dancing in the streets, because almost everyone else is stuck.
The other side of the table
The most useful stretch of the night was Devon being honest about fund math. A great fund returns about five times the capital its investors put in; ten times is crushing it, and under three is mediocre. The way you get there is brutal in its concentration: roughly five percent of the money has to come back a hundredfold, and almost nothing else moves the needle. Half of a VC’s investments fail, and a two or three times return, life-changing for the founder, is a rounding error to the fund. Paul Graham calls this black swan farming, and it means an investor’s real fear is not backing a company that dies; it is passing on the one that becomes a home run.
The honest consequence is that an investor never validates your company. They validate that you fit their thesis right now. That runs both ways: if an investor has gone quiet because your outcome no longer moves their fund, name it, be respectful of their time, and call them only at the moments they can genuinely help. Sunlight is a good disinfectant.
The economics underneath all of this have moved, and not subtly. When Garage started in 2015, the largest outcome anyone could point to was Instagram at a billion dollars, and that felt absurd. WhatsApp later went for nineteen billion and held the record for years. Groq took it last Christmas at twenty, then lost it to Google’s Wiz deal a few weeks later. SpaceX now sat near 1.7 trillion while still private. The point was not the trivia; it was that the prize in private markets is now a hundred to a thousand times larger than it was a decade ago, and companies can stay private long enough to capture it rather than handing it to public markets early. That shifts the math on almost every decision, including where you raise. Moving an entry valuation from ten million to twenty is immaterial when the outcome at the end is ten or a hundred times bigger.
Building today
So what does this mean for someone starting now? The leverage has changed. Devon described a founder doing the work of a full sales team essentially alone, using AI agents to run prospecting and outreach across a hundred fronts at once. That is the gift of the moment: a resourceful person can cover ground that used to take a headcount of dozens, and the prize at the end is larger than it has ever been. The hard part, the unglamorous go-to-market grind, is also the opening, because most people are waiting for it to get easier instead of doing it.
The catch is that the same tools are in everyone’s hands, which is why the durable edge keeps circling back to the founder. Build something a few people genuinely love, stay closer to the customer than to the investor, and be the kind of person who figures it out when the plan breaks. That is the part no model and no market shift replaces, and it is what Devon returned to all night.
Key Takeaways
The best pitch sounds like a stock tip among friends. One or two sentences on why this wins, and the rest is detail. If a founder cannot find that line, an investor usually cannot find it for them, and a long email is a tell that the work hasn’t been done yet.
Default dead can be a deliberate strategy. Companies building hard technology stay default dead for years by design, and that is fine when the eventual prize is enormous. It is not a licence for everyone else; most companies do not have that excuse, and should be doing the default-alive math instead.
Most software no longer has a technology moat. A weekend is enough to clone a lot of products, which pushes margins toward zero. The advantages that compound are customer-side: deep knowledge of the problem, being the system of record, and finding novel ways to reach people when attention is scarce.
You are a lottery ticket, and understanding that helps. Half of any good fund’s bets fail and a 2-3x outcome barely registers, even though it would change a founder’s life. Knowing you are one ticket among hundreds lets you manage the relationship like an adult rather than chasing approval.
Know your default-alive number before you decide anything about money. It is plain arithmetic on flat costs and current cash, and it is the thing that tells you whether to raise, how much, and when. Without it, every funding decision is a guess dressed up as a plan.
Don’t let geography set your ceiling. The prize in private markets is far larger than it was a decade ago, and capital now stays private long enough to capture it. The ambition you choose is the ambition you get priced on.
The S26 Founder Dinner Series is hosted by Builders Club in Waterloo. Sessions pair evening founder dinners with midday workshops. Thanks to our sponsors Osler, TD Innovation Partners, and Communitech for their continued support.