A note on what follows: As I take what’s been a five-year night-job full-time I’m putting pen to pixel in an attempt to articulate my lens on early-stage investing. I call this framework my own, though in truth it’s a synthesis of counsel from wise predecessors, observation of fantastic (and not-so-fantastic) entrepreneurs, industry literature (more history than blogs), personal experience — including more mis-steps than I would like to admit — and a splash of my own techno-optimism.
This is far from a formula. It is not meant to be an absolute calculus on predicting the future (I’ll leave that to Hari Seldon or the TechnoCore) but rather a lens through which to standardize the hundreds of pitches you see as an early-stage technology investor. Contrary to what many established VCs will have you believe, it’s not difficult to pick startups with prospects; investing is as much about avoiding the bad ones as identifying the good. While operational, financial, and technical knowledge are part and parcel to savvy early-stage investing, they have diminishing marginal returns. The highest contributor to portfolio success (read “IRR”) is quantity*quality of deal flow, and ability to get into the best. VC is an access game, and in access games the motto is location, location, and likability.
Criterion 1 of 2: Scored (100 of 120pts req’d for fiduciary-level conviction)
i.) Founder + Founding Team (40pts)
- Ability * Background: I look for uncanny founder-market fit, industry-familiar smart creatives who intimately understand a vertical but approach it in a novel way with new technology–e.g. Jeff Huber at Standard Cyborg, who’s an amputee building human augmentation software, or Jenna Brown at Shipamax, who became frustrated with the software in dry bulk shipping while at a commodities house. Unusual familiarity with a market is an easy, early defensibility-driver.
- Determination (Grit) * Passion (Conviction): i.e. Are they truly obsessed with solving the problem at hand? Why? What would make them give up? The entrepreneur’s motivation (or lack thereof) is a major risk factor. All seems fine and dandy when you close your seed round, but you know, Murphy’s Law.
- Honesty + Character (gut read, ‘reference checks’): i.e. Can I trust them? At Seed the entrepreneur has total control — as they should, though that’s another post. Not to mention, culture stems from leadership, especially the CEO (a certain Lyft competitor comes to mind). It is my firm belief that Silicon Valley has without regard for empiricism largely deified the wrong entrepreneurs (and investors, for that matter).
- Vision * Pragmatism (user-centricity): i.e. Can they inspire others to see a common vision while building a realistic bridge? It’s often easier to see the long-term than the mid.
- Salespersonship * Connectedness: i.e. Can they sell, and who will read their emails? Note: I believe the Sales Ability-to-Success curve is non-linear. That is, past a certain point it is dangerous to be better at selling as it can disconnect expectation from reality. I’ve seen a lot of vaporware shortsightedly sold to both customers and investors. Your product should largely sell itself.
- Team history & Cohesiveness: How do they know one another, and for how long? By what mechanism do co-founders resolve disagreement and how well do they move on? Co-founder disputes are one of a startup’s greatest risk factors. There’s no room for resentment on their journey, or for ego.
- Capacity for speed. A good measure of this is how many code commits you have per week.
- Bonus points for women, minorities, ‘disabilities’. Venture capitalists systemically undervalue these founders.
- Brand and culture are more important (for growth and defensibility) than most entrepreneurs realize. However, both largely extend from the CEO and are incredibly nuanced. I will judge your e-book by its homepage.
- I look for CEOs who hate raising capital, because they love building businesses that make money. Funding should be a function of traction on positive gross margins, not vice versa. If you’re spending more time fundraising than running your business, you’ve got a problem. Drop VCs who drag their feet — they’re not investing.
ii.) Technology * Defensibility / Competitive Landscape (35pts)
- (15pts:) UNDERLYING TECHNOLOGY: If you just hired a freelance developer to build you an MVP then your company probably isn’t for me. If your problem wasn’t hard to solve then you’ll have quite a few competitors who are likely at least as smart as you are.
- There are really only two ways to be long-term defensible: IP and Network Effects. If you don’t have that — and no one has network effects from the start — then a great team with a head start will buy you a year, maybe two (I saw this first-hand when Common Bond fast-followed SoFi). I’ll only bet on a ‘head start’ if there’s what I call “Competitive Osmosis” (see Note below).
- IP: Only acceptable if you’re truly ‘deep-tech’, but by then your defensibility likely derives more from a world-class team than from the patents themselves. I care not a lick for business process patents (they’re circumventable) and steeply discount software patents by “1,000 ways to code a cat”.
- Network Effects: When this term came into vogue a decade ago it referred to social network effects — i.e. each new user makes the network more valuable. This is why VCs like marketplaces so much: once the network is big enough then it’s very difficult to build a viable competitor. You wouldn’t start an Airbnb competitor today and network effects are why. But I’m more interested in a new type of network effect: ML/DL. Machine learning models improve, definitionally, with more data.
- I believe that any software company not leveraging machine learning (or that’s built as a distributed application vis-a-vis blockchain) has a very good chance of not being around in 5–10 years.
- (13pts:) Why now? A solution can easily be too late or too early. What technological, societal, regulatory, or market development prevented this company from being started three years ago? Note: Technologies can be too late or early too… IMO, AR/VR and drones are a bit too early (and over-hyped), while mobile is (way) too late.
- (3pts:) Brand (this is often under-valued, esp. by 1st-time entrepreneurs)
- (2pts:) Do they have killer advisers? I don’t mean your buddy who’s worked in the industry a few years more than you, I mean people who can pick up the phone and get you your first five customers or coach your engineers through technical problems no one has ever solved before.
- (2pt:) Cash on hand (Note: Cash-to-Defensibility is non-linear; too much cash breeds profligacy and complacency.)
- # of competitors (direct & possible; there’s always a competitor you don’t know about yet–maybe I’ve seen their deck)
- Agility of competitors
- Competitors’ cash on hand (over the “can catch up fast” threshold, discounted by solution’s prioritization on their product roadmap)
- As frustrating as it is, there are certain entrepreneurs and funds the lion’s share of VCs simply won’t bet against — e.g. Musk, Bezos, Sequoia.
Notes on Verticals & Competitive Osmosis (more on this in an upcoming post):
- First of all, I don’t invest in consumer technology. I don’t pretend to know what consumers want. While the largest technology companies are predominantly B2C, the competition is more fierce, defensibility more difficult, and the risk profile such that it requires a large fund that’s able to make many bets and follow on big in the winners. That doesn’t work with a small fund like Cantos. Companies are much more predictable. They only want to do two things: Increase top-line and reduce cost. Venture capital is a game of statistics, and I like my odds in B2B much better.
- Now, this is where vertical really matters. I look for sectors that have big, slow movers, and where your average Millennial entrepreneur isn’t yet focused (i.e. NOT social networking, ad-tech, marketing-tech, CRM/sales-tech, on-demand, e-comm, cloud infrastructure, dev tools). This could mean manufacturing (e.g. Bigfinite, Standard Cyborg), construction (e.g. BuildingConnected, ALICE), finance (e.g. Qwil, Maxwell), healthcare (e.g. Patch, GreenLight Medical), logistics (e.g. Shipamax, EasyPost), agriculture (e.g. Arable, AgriData) and other niche sectors that require knowledge thereof. Generally speaking…
- YES: Finance/insurance (blockchain solutions only; pure software solutions have been played out), manufacturing (esp. +AI), logistics, agriculture (though sensitive to robotics), construction (though ALICE may be my only bet here), gov/law (where ‘humans are serving as middleware’, to borrow a quote from Manan Mehta at Unshackled), synthetic biology (non-clinical), infrastructure for a decentralized web
- NO: Social, on-demand, ad/marketing/sales-tech, e-commerce, hospitality-tech, hiring tech (too many of you), AR/VR (too soon), bio-pharma (takes a whole other skill set), real estate (played out)
- Because I don’t/can’t know all of these sectors intimately, be prepared for me to ferret out a SME and revert with more specific questions.
iii.) Clarity of Business Model, Unit Economics (10pts)
- Business model simplicity (do one thing, do it well). One of my mantras is: Complexity is the enemy of speed. Sell one product to one demographic. Know your fixed and variable costs! Model in ~20-30% uncertainty.
- Day 1 revenue or <6-month path to product-related revenue. I typically invest in companies with $5–30K monthly revenue, even at ‘pre-seed’ (pre-money valuation under $5M).
- Gross-margins > 60%. While I wouldn’t echo his hallmark condescension, a quote from Don Valentine comes to mind: “We need guys [/girls] to have 70% gross margins, because they abuse so much money. You need to be in a business that has huge gross margins, because the managements are not very experienced. You start with a 60% minimum margin as the ideal, because you know the management is going to spend 20% on marketing and 20% on research. We have guys who are 27 years old and have never managed anything.”
- LTV/CAC > 300%. Payback period also needs to be taken into account–can’t be longer than your ability to fundraise!
- Churn < 3%. Past that you eventually churn through the entire market.
- The revenue model must be aligned to incentives of the customers. Note: Its violation of this principle is my fundamental problem with companies driven by ad revenue.
- Extreme bias toward low burn (sub-$50K/mo at pre-seed, $100K at seed). Use advisers for early sales, not FTEs, and don’t sign a lease before product-market fit!
- CEO will consider herself cash-poor until $5M+ raised. Until then use the currency you have: equity!
- Major preference toward enterprise — lower risk profile (granted, lower possible upside), lower churn, deeper pockets
- I will not invest in anything I feel is or could quickly become commoditized — UI-based products, storage, algorithms on publicly accessible datasets, cybersecurity, on-demand services (better called “labor arbitrage”), or anything that is ‘trending’.
- Advertising (eyeball-based) revenue is an unacceptable answer for my risk/return profile
iv.) Exit Opportunities / Valuation (10pts)
- Natural acquirers at multiple stages. I believe in optionality, not IPO-or-bust. Assume Google won’t buy you (though I do think their acq. activity will pick up; what else will they do with all that cash?)
- I want to know your market is large enough that a 10% share means at least $100M in revenue–$100M ARR roughly being the hurdle for a SaaS IPO
- Vertical has high M&A activity, incumbents have large cash reserves, industry is facing or will soon face sweeping changes at the hands of innovation
- Look for “arbitrage” opportunities, i.e. underpriced deals. Note: these are in fact fairly priced deals, often found outside of the Bay Area, where we’ve accustomed ourselves to price inflation.
- Bias against seed round >$10M post-money valuation, as most acquisitions happen under $50M
- Berkeley, MIT, & other engineering-heavy universities are overlooked opportunities. Stanford is a fantastic school but over-emphasized. Note: Wary of scientist CEOs unless innate business acumen is present.
- If your total addressable market (TAM) — and this should be real revenue potential, NOT GMV! — is under $2B, you’re wasting your time speaking with most VCs, particularly $100M+ funds. What types of outcomes VCs will bet on are primarily a function of their AUM & # of investment professionals.
- Acquisitions have slowed in recent months but Google, Apple, Facebook, Oracle, and others (e.g. Berkshire Hathaway’s $60B war chest) are sitting on huge piles of cash that they need to put to work. These cash piles will be heightened should the current administration incentivize repatriation of foreign cash. I believe Google & Facebook will become the Berkshire Hathaway’s of tech.
- I like the types of acquirers that are more likely to be in $BRK’s portfolio than Benchmark’s — e.g. GE, GM, Marmon, Honeywell, Mastercard, Verisk, John Deere
- Train thoroughbreds, don’t hunt unicorns. Nine-figure exits are OK for a small fund like Cantos, though I do need conviction you’ll be able to raise Series A.
- Seed is less affected by bear markets due to lower burn and valuation. As NEA Co-Founder Dick Kramlich once said, “Good ideas don’t go on strike.” Companies seed funded in ‘08-’09 include Airbnb, Twilio, Dropbox, Slack, Credit Karma.
v.) Financing Risk (5pts)
- Since I will write smaller checks, my co-investors must be able to carry the day. My $50–100K check will buy a seed-stage startup only 1–2 months…
- …Or the company must be riding a wave/in a market/run by a CEO that I’m confident up-stack VCs are likely to fund. Note: I know enough VCs to have a fairly good read on this ;)
- If the latter is the case, I want to be familiar with the metrics that will be required for a strong Series A and confident the CEO can hit them
- I want the round I’m participating in to buy the startup at least 12 months of runway. In this wary market, 18+ is preferable.
- Note: I always test companies against a few friendly VCs at larger funds before writing a check. I’m showing my hand here but if I introduce you to another investor I’m not being 100% altruistic. [Shout-out to some of my usual suspects… Arianna Simpson, Shivon Zilis, Zavain Dar, Vaughn Blake, John Komkov, Nan Li, Katherine Boyle, Ela Madej, Zal Bilimoria, Michael Berolzheimer, Trae Stephens, Anya Schiess, Mo.ElBibany, Howie Diamond, Alastair Trueger, Ryan Gembala, Seth Winterroth]
vi.) Additive Business (10pts)
The company’s product or service is “additive” to the world, i.e. contributes to:
- Health (e.g. Patch, Recursion, Enlitic)
- Financial Inclusion & Private Sector Accountability (e.g. Robinhood, Qwil, Ethic)
- Democratization & Public Sector Accountability (e.g. OpenGov, Vendor Registry)
- Environment, Waste Reduction, & Food Security (e.g. Opus 12, Tri Alpha Energy, Arable)
- Education (e.g. Duolingo, AltSchool, Treehouse)
- Quality of Worklife (e.g. Slack, WeWork, Ambition)
- Notes: I acknowledge this is highly subjective but hey, it’s my fund! This may read as an ‘impact investing’ filter but is in fact a financial decision. IRR stems from value-creation, and I believe the most valuable companies are created by solving the largest pain points. This is also a meaningful differentiator in hiring talented Millennials… as sometimes even getting away with under-paying them (until you can afford market rates).
vii.) Gut (5pts)
- The indescribable feeling that you should invest despite non-obvious reasons
- Note: I believe what we call “gut” is the brain’s own deep learning. It is a complex layering of so much information as to feel intuitive when it is in fact deeply logical. While it has an explicit 5% here, it affects all of the scores above.
Criterion 2 of 2: Binary
What risks am I taking, specifically (e.g. technical risk, market risk, funding risk, competitive risk, hiring risk, retention risk, co-founder fallout risk), and am I willing to take them as a fiduciary? I typically frame this by asking, What assumptions must hold true for this company to succeed? Then I attempt to give a % likelihood for each over 5- and 10-year timeframes.
Or, as one veteran VC put, “Ask yourself, What price glory?”