Why VCs sometimes push companies to burn too fast
blog.ycombinator.com
blog.ycombinator.com
> The company runs out of money.
Actually, the worst case is keeping a company going even though the fundamentals of the business are imprecise, hard-to-define or just plain wrong.
You end up wasting resources of all kinds for everyone involved. Investors, employees, and founders, to varying degrees, will squander time, energy, enthusiasm, money in pursuit of a mirage.
Money isn't everything, surprisingly, even in a round of funding. If you accept a huge sum of cash with a giant valuation attached, you must have reasonable certainty that you will at some point achieve that valuation without more raising. Mental gymnastics aside, I don't think most startups can guarantee outcomes, but they must be, at least, certain of their motives.
Are you raising because you're at the scaling stage? Or are you raising to buy yourself a little bit more time to hopefully, maybe, figure things out?
The latter reason is a poor reason to raise, and most of the time, only staves off the inevitable at the cost of unnecessary suffering and hardship.
A founder is a fool if they don't think the clock is ticking. Fools don't focus; fools fester.
Indeed, the article isn't even talking about putting more money into a worst case scenario; it's talking about founders aggressively using the resources they've been given.
Most 'employees' will never see any significant upside in any of the startups they've worked at or will work with - so their compensation is their livelihood and not exactly a waste of time.
There is a lot of inefficiency at big companies as well, so it's arguable that most tech workers are 'wasting at least some of their time' in the big disorderly march to progress.
They are also gaining skills.
But I agree it's a pretty bad thing to do to raise money to 'figure things out'. I'd loathe to think of how many of those would actually get funded.
If a company doesn't do well, the investor loses all/most of their investment. It doesn't matter how ambitiously or unambitiously the startup acted, the VC's losses are roughly fixed.
On the other hand, the difference between a company having a very good (e.g. 8x) vs. fantastic (e.g. 50x) outcome is huge. If increasing burn helps a company grow faster, acquire more market share, and improves the chances that the company will 50x instead of 8x, then that's worth a higher chance of failure to the VC because they have some diversification. It may or may not be worth it to the founders, who only work on one startup at a time.
So basically, if the two possible scenarios are the following..
1) 60% chance to 0x, 30% chance to 3x, 10% chance to 8x
2) 80% chance to 0x, 15% chance to 3x, 5% chance to 50x
.. then VCs strongly prefer the second scenario. There's a slightly higher chance of a loss, but the upside is much, much greater.
Investing in high growth models seems like a great idea when sitting around drinking beer, but I'm willing to bet the long term effect of this practice is going to come back and bite us in the ass sooner than later, if it hasn't already.
There's a reason why seaweed reduces cow gas. We have to use seaweed on cows now because we removed the cows from the grass. There's a similar reason behind why our privacy has been thrown out with the bathwater!
1) 80% 1x, 20% 5x (EV: 1.8x)
2) 99% 0x, 1% 100x (EV: 1.0x)
How many VCs do you know who would take 1 over 2, even though it's the "obvious" money-winning bet?
I'd love your thoughts on why this is. My semi-cynical thesis is that VCs need home-run logos on the Portfolio page of their website to keep their spot at the dealflow table.
Let's break the VC world up into three types:
A - Brand-name VCs that funded yesterday's massive unicorns, and are regarded as clairvoyant masters of the universe
B - Unknown VCs with a half-decent return on their last fund, but didn't back anyone you've heard of
C - VCs that swung for the fences a few too many times and imploded
Founders want to work with "A" VCs (for both rational and irrational reasons), so the first names they call are the ones whose past big bets paid off. "B" VCs are much lower on the call list, and only get access to the deals that "A" VCs all passed on. "C" VCs are obviously not available to chat.
It is arguably rational for a VC to seek to become an "A" and end up as a "C", rather than to muddle along as a "B" until she's getting pitched every week by hair salon and gym owners with A Great App Idea and her LPs wonder what she's doing with their money.
First, I don't think any rational VC would take #2 over #1 in your example, but the caveats are: a) people aren't good at estimating small probabilities, so they might think that #2 should be 97%/3% instead of 99%/1%, and then #2 is "better". And b) They overestimate how much of a "sure thing" something is, so maybe what seems like 80%/20% in scenario #1 is really 50% 0x, 30% 1x, and 20% 5x. I can't remember the origin of the quote, but I've repeatedly heard that "trying to build a $100m company is not that much riskier than trying to build a $10b company, so you may as well shoot for $10b."
Also, generally VCs won't take bets where the EV is <3x or so. A good fund will have a net return of 3x+, and you can't get there by making a lot of 2x bets.
Furthermore, 99% 0x vs. 1% 100x is very lopsided. Probably too lopsided for most investors with portfolios of 20-40 companies per fund. If someone made that bet for all 30 companies in their fund, the fund would be 0x almost 3/4 of the time. But if the bet was more like 95% 0x and 5% 100x, then that's kind of interesting. Making 30 investments like that would produce a terrible return 21% of the time (0 wins), a good return 34% of the time (1 win), and a terrific return the rest of the time (2+ wins).
That said, prestige is (unfortunately) a factor. So I could imagine someone making an investment with an EV of 4x that is likely to be a unicorn over an investment with an EV of 5x that is unlikely to be a unicorn.
The last thing I'll add is that A/B/C grades for investors are context-specific. An 'A' like Sequoia or Accel might be generally pretty great for everyone, and you won't go wrong taking money from those firms. But in certain areas, lesser-known funds might be very strong company-building partners. For example, there are specialty funds that are focused on SaaS companies or frontier tech, and if you operate in one of those spaces then those funds might be just as good as taking money from Sequoia.
Finally, below is an interesting set of tweets from an LP who has been investing in VC funds for several decades. TLDR: it's basically impossible to build a great fund return on top of lots of small wins; you need a few huge wins instead.
https://twitter.com/HorsleyBridge/status/657287940456886272
https://twitter.com/HorsleyBridge/status/657288063329013760
In other industries like biotech it's not that rare to see early rounds where founders give up 40-50% of the company so even though investments are bigger and riskier, the financials work out to returns as good or better than tech in general. The pharmaceutical industry has outsourced large amounts of R&D risk to biotech investors for a long time though, so the risk is usually scientific and not product-market fit or sales/distribution. If you're working on a drug for a certain disease or condition, it's pretty easy to calculate how much you can make or how much you'd be worth to a pharmaceutical giant at various stages of risk (seed, animal studies, efficacy studies, clinical trials, post approval, etc.).
The causality probably isn't:
LPs want home runs -> LPs look for home runs on VC pages -> VCs deliver home runs.
Rather, it's probably:
In the VC space, home runs make up the majority of the EV -> This common fact supports LPs looking for home runs, and VCs looking for home runs.
There are probably some synergistic effects where the two are more than additive, but I think the underlying cause must be that in VC space, home runs make up the bulk of the return.
To see this, consider PE firms or mutual funds -- are these industries a lot less into home runs? I would argue yes -- that for these firms they are measured by LPs by their historical average returns, or fees, or whatnot. Because in these firms that's the closest proxy for skill. Whereas in the VC world, home runs are the proxy for returns.
(The author dances around this, for understandable reasons.)
This is actually something you can use when negotiating if someone tells you "I need 20%." Really, they need a big return on their invested dollars. The exact % is just a rule of thumb based on the overall returns they expect from their investments in the context of the entire portfolio.
Investors manage portfolios and fund returns are dominated by the home runs.
Therefore, investors generally want to maximize risk (and potential return), which is usually at odds with a founding team with most of their net worth tied up in the company.
Let's say you invest 5-10 years into a startup. Which would you rather have? A 30% chance at a $10M payday, or a 5% chance at a $100M payday? The answer is not "Well, I'll just repeat until the 5% chance materializes."
For example, if a company has an 80% chance of getting sold for at least $50M, the founder wants to make sure that happens. A VC is more interested in the 10% chance that the company could sell for $200M and the 2% chance that it will sell for $1B.
A faster burn rate doesn't materially impact the VC's odds of "success," because companies destined to be unicorns won't be hurt by a faster burn rate. But burning cash makes it much more likely that the company is sold for scrap instead of hitting the 80% chance of a $50M exit (which means life-changing wealth for most founders).
I don't think it would be a horrible idea for some of the bigger funds to set up a separate "race to profitability" fund where the explicit goal of the companies is to be profitable as soon as possible. You could even hold on to the winners from this fund long term instead of selling your equity share and aiming for the magic exit and infuse the generated profits into the fund. Could be a great learning exercise and also a decent instrument to have for the normal fund. I can very well see a "hard talk" where VCs tell one of their normal portfolio companies...look we think your basically a bust and will transfer you to the other fund where you have N time to become profitable on a smaller scale. Thankfully we have some experts there who can trim the hyper growth engine off your company and help you focus on smaller level profitability.
Would really like to see such a list, and be able to pick and run it further?
The thing is, the CEO has started his own company twice and sold both (this was his third company) so we both assume he knows what he's doing.
In the scenario of the CEO, I feel like his plan is this, and only this:
"high risk, high reward"
"live fast, die young"
Are these the correct thoughts for a serial entrepreneur (who is also sorta like an investor)?
*edit, funding was actually for series A not B
It seems that newer entrepreneurs will take less risk (is this true..? that's why I used seems). More seasoned/serial entrepreneurs will just go balls to the wall because they know of the mistakes they've made in the past and know how to be successful. They don't seem to be as scared of failure because they've failed before.
Should newer entrepreneurs just skip the progression and go straight to not being afraid and effect mimic more seasoned peers. Would the chances of success be slightly higher (even a slight higher chance of success is worth it...)?
When a VC pushes for a faster burn which then necessitates a new VC firm to fund the next round (perhaps prematurely relative to the founder's perspective), it's so that that individual VC can mark up their personal IRR.
Which may be relevant to their firm and the firm's fundraising plans to LPs. But could also be even more relevant if that partner is thinking about changing firms or starting his/her fund own. Individual IRR, based on deals with attribution, is the resume of a VC.
The effect of this is, junior partners push for this type of behavior more than senior partners that are already set with an impressive personal IRR.
Problem: Using just traction and its rate of growth is a poor way to evaluate the real future of the company.
Here's some of how: That approach ignores (A) the size of the market and its growth rate, (B) the core technology and is it defensible or, instead, easy to duplicate or equal, (C) the status of would be competition, (D) the scalability of the technology, (E) the margins once have reached some significant scale, etc.
The VCs look like they would really like a drag race car that from a standing start just went 0-100 MPH in 2.5 seconds but 500 feet ahead is headed off a cliff, in another 1/2 second will have the rear tires explode, the supercharger belt fail, or already has the engine at 8000 RPM and in 1/2 second at 10,000 RPM the engine will explode.
Why Do VCs Do This: As the OP explains, the VCs have to report back to their LPs quite frequently, and the LPs are looking only for simplistic measures, e.g., from something close to traditional accounting, e.g., would ignore anything about the engineering of a race car.
Apparently the VCs and LPs would prefer a lemonade stand that opened just before July 4th and on July 5th can show good traction and fantastic rate of growth in traction.
For a safe, effective, cheap one pill that taken once can cure any cancer now in Phase III trials, the information technology VCs would say "No traction" and f'get about it.
So, the VCs want an early pop easy to see and measure in simplistic terms. Okay, that can become a game. So, delay all reports of progress. When have accumulated some nice progress in traction, in January say that will have rapid growth starting in March and will have a report on April 1st. Do that, and have the April 1st report look really good on traction and its rate of growth. Then let the VCs report this fantastic exploding success back to the LPs.
Information technology VCs just flatly refuse to evaluate projects based on the real internals and, instead, pay attention only to simplistic, visible, external measures. That's foolish.
Entrepreneurs beware: Reporting to a BoD of fools is no fun.
On biotech, I had a sailing buddy, Cambridge PhD, and I could not understand the motivation of principle engineer level researchers in that field. Maybe it's a job job; I couldn't see any parallel.
So, a guess is that with such backgrounds, the bio-tech VCs can look at a bio-tech startup and look at its technology and actually understand and evaluate it, understand the competitive situation, look at market potential, evaluate the technical qualifications of the founders, etc.
E.g., for bio-tech projects, the NIH problem sponsors can do such evaluations. For STEM projects, the NSF, DARPA, ONR, etc. problem sponsors can do such evaluations.
The corresponding situation for information technology VCs is wildly different -- history majors who got MBA degrees and have some experience in business development.
With high irony, it was leading edge STEM field technical work for US national security and the space race that initially got Silicon Valley going -- transistors, microwave, radar, etc.
Two major reasons here seem to be 1) VCs not wanting to sink time into a failing startup (force downside) and 2) markups in valuation are preferred by VCs (force upside).
But directionally these incentives are the same for founders. Founders probably want to work less if their startup is failing (per person*years they put in). Founders also prefer the case where they're valued a lot and worth a lot.
Magnitude-wise these motivations are very different between founders and investors. For founders, a failure means lost years of their lives -- forcing a downside is a lot more painful. Founders also celebrate the home-run as much as the VCs (in fact probably for the biggest home runs, the founders get out more than an individual VC right?).
This could create an asymmetry whereby founders are in no hurry if hurry could mean failure, and the model above is classical risk aversion. Conversely, VCs don't mind that much if your chance of failure goes up by a few percentage points.
Aaron also tacks on some new ingredients: VCs like to see markups for it's own sake (to solve the monitoring problem between LPs and VCs).
I have no doubt with his experience the mechanisms are absolutely correct, just wondering if the majority of the underlying generator of these symptoms is underlying risk aversion.
The only difficulty would be the first couple funds where you'd have to push group 1 to exit so you could raise fund 2, then try to raise fund 3 on the exits of fund 1 plus the growth (but not exits) of fund 2.
If you look at the portfolio of a VC fund you'll see why. If they're in 45 companies, essentially all the return will come from 3-5 companies. The other 40 board seats are a hassle. Whether they return 0x or 2x, they're just not material to the success of the fund.
It seems like there could even be a market for taking the ok but never great companies off the hands of boom/bust mentality funds at fairly good conditions.