The 3-Question Test I Use Before Putting a Dollar Into Any CompanyVIEW IN BROWSER | FOLLOW LUKE ON 𝕏 
Hundreds of business plans, which their founders worked day and night on, mostly died on a single desk in Boston. It was Georges Doriot’s desk, a Harvard Business School professor with an eye for immaculate suits and unproven companies. Decades before anyone coined the phrase “venture capital,” Doriot was already putting it in motion, one rejection at a time. His students remembered one rule above all the others, because Doriot repeated it constantly: consider backing a grade-A founder with a grade-B idea. Never back a grade-B founder with a grade-A idea. That rule earned its keep in 1957, when a 31-year-old MIT engineer named Ken Olsen came looking for capital to build computers. The idea spooked the money men of the era, as a string of expensive corporate failures made “computer" practically a curse word. So Doriot’s team advised Olsen to drop it from the pitch and lead with the products instead. The company became Digital Equipment Corporation. Doriot’s firm, American Research and Development, paid $70,000 for a 70% stake. By the time that stake was finally distributed to shareholders, it was worth more than $350 million. That’s on the order of a five-thousand-fold return, and the trade proved venture capital could work. Doriot saw thousands of proposals. Of those, he funded a sliver. But his fortune came from that filtered approach. Over the past week, I’ve shown you how this era’s biggest fortunes are forming – Spark Capital’s early stake in Anthropic, the Roth IRA built on pre-IPO shares of Roblox Corp. (RBLX), the door Genentech’s IPO opened. And I told you the velvet rope that kept ordinary investors out of these deals is finally moving. I also told you access is only the starting point. Today’s issue is not another story about a fortune someone else already made, but with the actual mechanism I use before a single dollar of my own moves into a private deal. Three questions. The same three, asked in the same order, every time. I’m going to walk you through each one in enough depth that you could run it yourself on the next deal that crosses your desk, long after this issue is closed – who’s actually running the company, whether the thing they’ve built solves a problem people can’t walk away from, and whether the clock on the specific opportunity and the clock on the broader market both happen to be pointing the same direction at once. The Number Behind Every Great Venture StoryI need to give you an uncomfortable number before we go any further. Roughly nine out of ten early-stage bets fail to return their investors’ money. Across a typical portfolio, something like one in ten produces the outsized winner that makes the whole portfolio work. And more than half of the positions you hold will likely hand back less than you put in. That is the ordinary, expected math of investing in early-stage companies, even for professionals who do this for a living. Walk through a newly opened door and put money into the first company you see, and you’ve simply found a faster route into that “nine out of ten.” The door gets you into the room. The framework is what keeps you solvent once you’re inside. So today I want to hand you the actual framework. This is the same three-part framework I use on every company before I ever recommend one – to members, or to myself. I call it PPT: People, Product, Timing.
Recommended Link |
|
Eric Fry says yesterday’s AI sell-off was only a warning shot. He’s seen this late-stage bubble setup before — and it ends brutally for investors who aren’t prepared. He recorded a free video explaining what’s coming next and what to own instead of big AI stocks. Get the “Survivors List” at no cost here. |
 |
|
People: The Filter That Matters MostIf I had to weight the three, I’d honestly call it closer to People, People, People. Or, Doriot’s grade-A rule, rebuilt for 2026. You can find the perfect product launched at the perfect moment, and it will still fail if the people running it can’t execute under pressure, adapt when their first plan doesn’t work, or hold a team together through the inevitable stretch where nothing is going right. Arthur Rock followed this framework when he backed the “traitorous eight” young engineers who fled William Shockley’s lab – a story I shared with you a few days ago. Rock had no product to evaluate and no revenue to model. He bet on the people, and the bet became Fairchild Semiconductor, the company that seeded almost everything we now call Silicon Valley. The tricky part is that the “right” founder rarely looks like what you’d expect on paper. Take Tobi Lütke, who co-founded Shopify Inc. (SHOP). He left traditional schooling in Germany after the tenth grade to enter a hands-on computer programming apprenticeship, having already taught himself to code by rewriting his own video games around age eleven. On a resume, “left school at sixteen” can look like a red flag. In reality, he’d been doing the actual work – writing software daily – for longer than most of his eventual competitors had been alive. Lütke and his co-founders set out in 2004 to open an online snowboard shop. The existing e-commerce tools frustrated him so badly that he built his own platform to run the store. That side project became Shopify. And the lesson generalizes well past dropouts: the signal you actually want – deep, obsessive competence in the specific problem the company is solving – often never shows up on a conventional resume. You have to look past the credential and evaluate the person. Doriot would have called Lütke a grade-A founder on sight. In practice, that means checking the boring stuff first: what these people actually built before, where their claims can be independently verified, and whether people who’ve worked with them – other investors, other founders, former colleagues – describe them the way the founders describe themselves. Most candidates fail that process. Product: Does It Solve a Problem People Can’t Live Without?Once the people check out, the next question is simpler than most investors make it: does this product meet demand so urgent that customers will tolerate an imperfect early version just to get access to it? Google is the textbook case. By the late 1990s, the web had exploded into more information than anyone could organize, and search became the only sane way to navigate what had just been built. That urgency is why Andy Bechtolsheim wrote his famous $100,000 check on the spot – made out to a company that hadn’t even been incorporated yet. Early investors who saw the demand clearly were rewarded on a scale that still gets cited in venture circles decades later. I bring this up now because I think we’re in the early innings of an equally large unmet-demand story: physical AI – robots that can do real work in the real world. But the true chokepoint of physical AI sits in the training data. A language model can learn from the text of the entire internet, because that text already exists. A robot has no equivalent shortcut – every single skill has to be taught through physical demonstration, recorded frame by frame, often with a human operator physically guiding the machine through the task in real time. Industry data on this has moved fast: the cost of collecting a single hour of usable, labeled training data has fallen sharply over the past two years thanks to better tooling, but it remains a meaningfully expensive, physically constrained process – and a robot that needs to handle the real world’s endless variation reliably needs enormous volumes of that data, far beyond a few hundred hours of demonstrations. Whichever companies figure out how to produce that training data cheaply, at scale, are sitting on what I think becomes the defining chokepoint of the physical AI era – the equivalent of what chip design and cloud infrastructure became for the last decade of AI. That’s the “insatiable demand” test, applied to where I think the next major AI buildout is actually headed. Timing: Two Clocks, Not OneThe last piece of the framework covers two separate things people often collapse into one. The first clock is your entry into a specific company. It’s early enough that you’re buying in before the bulk of the value has been captured by earlier rounds. But it’s late enough that the company has proven its basic idea works. That balance is exactly why the People and Product questions come first; allowing you to judge whether a company at an early stage is actually derisked enough to be worth the timing bet. Remember Genentech: the investors who rushed through Wall Street’s door in 1980 were buying pure possibility. The ones who understood what Swanson and Boyer had already proven in the lab were making a judgment call. The second clock is the broader environment you’re investing into. And on that front, I’d argue current conditions are about as favorable as they’ve been in a long time. I walked you through the arithmetic a few days ago: the largest technology companies are pouring roughly $700 billion into capital projects this year. About $2 billion every single day… And they’re competing with each other for capability itself. Companies with that much money and that little time tend to buy what they cannot build fast enough. That competitive pressure compresses the timeline between “promising private startup” and “life-changing exit for early investors,” whether the exit comes through an IPO or an acquisition by a giant racing to secure capability it doesn’t have time to develop from scratch. Increasingly, the buyout is becoming the finish line early investors aim for. What This Framework Actually Buys YouI want to end where Doriot would have ended: nine out of ten of these bets still won’t work out, and no framework – mine or anyone else’s – changes that basic arithmetic. Even Bessemer Venture Partners, one of the oldest firms in the business, keeps a public “Anti-Portfolio” of the winners it passed on – Google, Apple, Airbnb, FedEx. Anyone who promises you a system that turns venture investing into a sure thing is lying to you. What People, Product, Timing actually does is shift your odds. It won’t turn a bad bet into a good one. But applied consistently, it filters out the startups that were never going to make it past the “nine” – the founders who don’t hold up under scrutiny, the products solving problems nobody urgently has, the companies you’re either too late or too early to matter for. What’s left is a much smaller list, and it’s the list worth staking capital against. Doriot read thousands of plans to find one Digital Equipment. However, you won’t need to read thousands. On July 30 at 1 p.m. Eastern, I’m going to walk through this exact framework live, in real time, applied to a specific AI company I’m recommending for free. You’ll see exactly how the People, Product, and Timing questions play out against a real deal, with a real opportunity to act on it the same day… plus two additional recommendations reserved exclusively for members who join that day. Reserve your spot for The 2026 AI Megadeal Event Sincerely, |
No comments:
Post a Comment