How to Raise Money (2013)
paulgraham.com
paulgraham.com
YC itself is opaque. All their rejection emails are a simple "sorry, we can't tell you". I don't know if they are afraid of legal liabilities or the reasons are much more mundane : "It's been 2 years and your revenue is only $2k", "AI is overhyped", "Too small a market", "You look dodgy" etc.
>Don't raise money unless you want it and it wants you.
I wish I knew this before I went on a wild goose chase for angels.
The only two startup articles I read are either PG or indiehackers and I am impressed with how clear and lucid this dude is with his words.
That's not at all what raising from VC firms is like. VCs rarely tell companies "no"; in fact, they say the opposite: "we'd love to fund you, let's keep moving down the path to us funding you, here are some next steps". They will do this indefinitely, and continue long after the decision not to fund you has been made.
That's what Paul Graham is talking about.
If someone wants to date you or hire you, they will make an effort. They will respond quickly. They make genuine counter-offers ("How about Monday instead?").
If they are lukewarm, they will string you along and eventually fade. They will never make counter-offers, just reject your offers and let you make the next one ("I can't on Tuesday").
If they are not interested, they won't reply in the first place.
When someone wants you it's obvious. And when they don't, it's obvious by omission.
It's sad that it took me so much time and heartache to work this out.
Think about it from the other side: if you were seriously considering making a company one of the few per year that you give millions of dollars, would you ignore their emails or cancel meetings at the last minute or sound bored on the phone? Investors get just as excited about a company they like as founders do about an interested investor or potential big customer.
But VCs keep stringing people along because they don't want to miss out at some point in the future if the business takes off.
That's different to dating or hiring (generally speaking).
They absolutely do not.
We found out 3 hours after the 10pm time they stated. And apparently this happens every year.
This is actually quite similar to your description of other VCs.
I mean, if you had to describe how a standard VC firm with the one difference is that they work in batches would reject someone, this is exactly what you'd predict - not right now, but feel free to apply again later. You can even apply in between batches although this isn't encouraged as much. They don't ever want to discourage you from giving them another shot to invest.
Nobody is expecting VC firms to say "no, and never reach out to us with this idea again". Maybe you haven't had the experience of raising from VCs before, so I'll relate it to you: they mostly say yes, and actually put some effort into making you believe they're actively interested, right now. They aren't. It is very difficult to know where you stand with them (right now), and they are more than willing to waste your time.
But why? What is the motive for doing so, if they don't want you?
Maybe things have changed in the last couple years, but for first raises for companies, this wasn't like, "a thing that happens sometimes"; it is the normal, standard behavior of VC firms. People who are good at raising have strategies for dealing with it.
Timing is a big factor in tech companies. Ideas which didn't work in 2010 can work now.
Also you can come back with more proof of your thesis.
Just about my favorite thing about good business writers is that, just like in a good sale or pitch, they communicate clearly, concisely, and with a serious economy of words.
Conversely, never trust a business person who can't write a simple article.
But this is pretty much the whole point and most startups DO need money because without it nobody can hire anyone.
Often times you even need money to build your product in the first place.
Some of them can make revenue from the get-go, some can't. But the truth is most of them would not have survived this long without first round of seed money.
The context is that when you're in Silicon Valley, the upside is you have lots of people that understand what you're trying to do, and a support network that helps you do what you're afraid to admit to yourself that you're capable of. The downside is it's easy to do what all your friends are doing.
If you see your friends raising, you might feel compelled to raise as well. Some people jump in without really thinking it through.
Reality is, money comes with strings and expectations. Too many people raise without understanding what that is, and suffer the consequences when they put their sweat and tears into their company.
For a newb, it just sounds like quite a few things to remember, lots of expensive mistakes to make, although the saving grace is that investors are about money so even if you call them a rude word, I'm sure they'll come back if you're making enough money!
Does anyone have a fixed link for the Sequoia slide deck?
because i feel like that is the best alternative to this 2013 article, it lets you bypass the whole problem of having to know an investor, mind you, is not necessarily a rewarding experience.
the fact is any one that has managed a small portfolio with or without wins is tough to tell apart from each other. when you have certain amount of money it buys you a comfortable cushion to fall back on despite your failures and still come across as successful, but in fact you are a horrible fucking money manager/investor who happened to get lucky in the 90s through ultimately being exposed to an environment where wealth and poverty divide is particularly wide in a massive city.
a crowd source equity fund raise, not much different from an ICO apart the added rigor (one ICO had some Russian dude in a drag posing as HR, switching gender and hair color for different roles filled by the same person, exit scammed with an anti-semetic pepe meme), equalizes the power balance between investor and innovator from what was a master-slave relationship to many masters-slave that is supposed to be more democratic yet fragility is questionable. ex. FUD, fake news causing run on bank vs one investor weathering it out because of #faith
im still not sure what the legality of crowd equity raising is in canada but this is what i decided to pursue now.
> Though it sounds slightly paradoxical, if you want to raise money, the best thing you can do is get yourself to the point where you don't need to.
This is akin to being unable to get a credit card when you Really need the money. Viewing it in these terms, it actually makes a lot of sense.
It's as though investors don't think at all, they're just following each other and copying every move. That's the definition of dumb money.
now the music is stopping and there's not enough chairs.
But how can you get introduced to them when they won't talk to you?
My recommendation is to get into incubator or even "lift off" programs and then work your way up.
The current startup culture makes it impossible for us to find "Bill Gates" in Africa/Asia because they have no visibility, no voice and no money and not because they are less talented
I believe the next cohort is early next year.
I'm working on a new social document collaboration platform:
and I'd like to raise funds for it but it's REALLY unclear if I should go ICO, pre-seed, etc.
Why deal with any of that headache when you can just launch a landing page/white paper and ask for money from the public to fund your idea without giving up equity or control.
There is a difference between regulated and illegal.
The way I read that is there are benefits. Seems to me it’s significantly more cost effective than a private offering and order of magnitude more cost effective than a traditional IPO.
>and some huge regularity drawbacks (not knowing who owns the shares).
Not knowing who owns the shares is in fact illegal.
Before 9/11 many states allowed it (as did many countries)...now no States allow it and as far as I’m aware Panama was the last country to outlaw the practice.
But if you know of a jurisdiction let me know...
I was under the impression ICOs could now only target accredited investors. In that sense it's really no different from raising capital the traditional way.
But the US is awash, from border to border, in successful businesses, say, Main Street businesses, that never got external funding. They were, say, bootstrapped.
Maybe if pick an information technology (IT) project carefully, then can bootstrap it or start it with just own checkbook and no external funding.
Really, an IT startup should need less money to start than a lot of Main Street businesses. E.g., can quickly need $100,000+ in spending to get to the first $1 of revenue in an auto body shop, auto repair shop, asphalt driveway paving company, dentist's office, pizza or Chinese carryout restaurant, etc. Heck, in my neighborhood, the guys mowing grass show up with $100,000+ in equipment.
In contrast, for, say, a Web site, can plug together a quite powerful server for less than $2000 and can get a lot of relevant software tools and infrastructure software for free or nearly so. Can run ads from ad networks. If people like the site and keep a $2000 server busy, then in a year can have more in pre-tax earnings than could get from a venture seed round.
Another counterexample -- Drudge Report.
Main Point: A carefully conceived and planned sole, solo founder information technology startup, e.g., a Web site, can be cheaper to start than even a common Main Street business, e.g., a pizza carryout, even a grass mowing service. But if the site becomes popular, then quickly it will generate more cash per year than in a seed round. From there the site might continue to be a significant business, e.g., Plenty of Fish, Drudge Report, and likely some more, still without equity funding.
Conclusion: The Web is one heck of an opportunity for startups without equity funding.
There is a common assumption in your argument, that the startup must have a period of wildly rapid growth to take the opportunity or lose out. Sometimes this is true, but as in my counterexamples and, really, just common sense, it does not have to be true and isn't always true.
Part of this argument about rapid growth and need for equity funding is that once a startup shows promise, it will be copied by others, and only a startup that executes well with lots of equity funding can win. This situation need not always be true: Instead, some startups can have crucial, core, powerful, valuable, proprietary, protected technology, secret sauce, locked up in a secure server farm, and difficult to duplicate or equal. That secret sauce can provide a powerful barrier to entry, "Buffett moat".
For my startup, I am a sole, solo founder. I've done all the work, have the software apparently running as intended, and am rushing to go live on the Internet.
From the timings of my software and some Mary Meeker data on ad revenue, the one server I plugged together, for less than $2000, if users like the site well enough to keep the server busy, should generate top line revenue and, essentially the same, pre-tax earnings of about $250,000 a month. I don't need the equity funding now, and with $250,000 a month in revenue, I will not need and would not accept equity funding.
My Web site is intended to be of high interest to essentially everyone in the world with a standard Web browser.
If my site gets to $250,000 a month in revenue, then that will be a significant sample of what users like from which we could make a projection to rapid growth to about $10 million a month, all just from my spare bedrooms and present Internet bandwidth. With such revenue per server dollar, the costs of the servers, etc. are next to trivial.
Thus, from pleasing a large fraction of everyone on the Internet, we could reasonably extrapolate the potential of my startup to a company worth $T+.
We have to keep in mind: Successful information technology startups are unusual things, each one essentially a unique business. So, we have to expect, at least not reject, some unique aspects.
Moreover, we have to see that the most successful startups were quite different from anything that went before. So being new and different, even radically different, are not sufficient for success but are nearly necessary -- again, we must not reject things just because they are new, different, unique.
Some products and services have such coveted utility that customers, users are really eager. My startup is so intended. Uh, my startup is squeaky clean, fully "safe for work", actually, say, culturally uplifting, an aid to better lives, families, communities, countries, and civilization, nothing to be ashamed of.
While my startup is a Web site, really what's important about it is just some original applied math I derived. The math is difficult to duplicate or equal because it is clever and original and exploits some advanced pure math prerequisites. Users will just get the utility and not be aware of the math.
But it is the math that is the crucial core secret sauce, difficult to duplicate or equal. In comparison, except for the crucial core, the software is next to trivial. So the site is not really a computer software development effort but an applied math effort.
There are not many people who know the math prerequisites and can do good original applied math, and the people that can nearly never want to do such applied math, want nothing to do with software, and are not in the information technology startup community.
There is a way to have a good idea about the promise of my work -- based on the utility for all the people on the Internet with just a standard Web browser: While Silicon Valley does not use such means of startup evaluation, several very important parts of our civilization do, do so routinely and quite successfully, with batting average far, far above that of Sand Hill Road.
So some parts of our society can look at a serious problem, see clearly that the problem is important and the first good or much better defensible solution will be very valuable, and evaluate proposed solutions with high reliability. Such parts include some internal industrial R&D, Ph.D. committees, editors of the better STEM field peer reviewed journals of original research, NSF, NIH, ONR, DARPA, USAF Cambridge, NASA, major engineering firms, etc.
My work would easily pass such filtering. So, I'm confident.
In strong contrast, Sand Hill Road wants to see a minimum viable product with traction significant and growing rapidly; all the project evaluation techniques of Ph.D. committees, NSF, ... are ignored and set aside with severe determination.
Sand Hill Road is not interested in anything from original STEM field research, maybe is forbidden by their limited partners to consider such research, and is not practiced in evaluating such research.
But I hold a good Ph.D. in pure and applied math from one of the world's best research universities. All my career, many times, I've successfully applied math, sometimes original, to practical problems.
So, I have a solid basis on which to be optimistic. Sand Hill Road can't understand that.
Sand Hill Road wants nothing to do with me before my business is so good I will want nothing to do with them.
Yes, my startup is unique, and that's nearly necessary for major success.
It would be short sighted to reject what I'm saying merely because it is unique or even different.
On crushing it, that's not part of my intentions or planning. The $1T+ is.
I've seen a lot in computing, e.g., in an AI, artificial intelligence, project at IBM's Yorktown Heights Watson lab, for US national security, saving FedEx twice, teaching ugrad computer science at Georgetown University and graduate computer science at Ohio State University, and I've seen much more in pure and applied math.
Net, at this point, my considered option is (1) academic computer science has run out of promising research directions they have the qualifications to pursue and (2) Sand Hill Road style computing is essentially dead and all but buried. Routine software will continue, much as routine sheet steel stamping has continued, but for now the glory days are over. In particular, for AI, etc., I see 99 44/100% hype and nearly all the rest water or mostly quite old applied math. We just had an AI spring of hope then a summer of hype and are now in a fall of failure and soon again in an AI winter. E.g., to me self-driving cars in current traffic on current roads is obviously an absurdity and a lot of hype, a bubble about to burst.
Computing remains a major opportunity, but nearly none of the software development community, academic computer science, or Sand Hill Road know how to make use of the opportunity.
For exploiting the amazing cycles, bytes, bandwidth, infrastructure software, etc. some new directions are needed.
IMHO the most promising such new direction is engineering based on very carefully done theorems and proofs of pure/applied math.
The number of people in the world ready, willing, able, and eager to do this could have a convention in a $20 a night motel room.
So the situation is bad, but the flip side is a grand opportunity. E.g., the processor I bought is an AMD FX-8350, 64 bit addressing, 8 cores, 4.0 GHz standard clock speed, now available quantity one, retail for $100-. ECC, error correcting coding, main memory is available for $10-/GB. The situation on hard disks, solid state disks, and Internet bandwidth is maybe even more amazing. So, the need and the opportunity is to make good use of such products.
First step: To heck with Kernighan and Ritchie, Stroustrup, and even with Knuth. Instead get a copy of W. Rudin, Principles of Mathematical Analysis, Third Edition, dig in, and work nearly all the exercises.
E.g., again, once again, over again, yet again, one more time, US national security is going deep long ball, if you prefer, swinging for the fences, into challenging pure and applied math with the new Space Force. For the role of math there, trust me!
No, don't trust me: Instead watch the start of the movie about John Nash where the flat statement is that "mathematics won WWII". Exercise: Discuss with arguments pro and con. Don't forget that Commander Rochefort's work and advice to Admiral Nimitz is what sank four Japanese carriers at Midway and turned the tide of battle in WWII in the Pacific. Uh, the Japanese brought 6 carriers to Pearl Harbor. Soon they lost one at Coral Sea -- also due to Rochefort. Then they lost 4 more at Midway. So early in the war, they were already running short of aircraft carriers. And don't forget much of the reason Rommel lost in North Africa was because most of his supplies went to the bottom of the Mediterranean due, right to code breaking. And note the importance of the cavity magnetron from Maxwell's equations in winning the Battle of Britain. And there was much more, e.g., from A. Turing. Then we have to count The Bomb, e.g., the critical mass calculations.
If you regard these views as new or different, then just remember you first learned them here.
Whenever I found a way to apply research-level mathematics to a business problem, the solution always was too difficult to sell. Sometimes I can solve my own problem, e.g. options pricing models known from financial mathematics can tell you the minimum to accept in a salary negotiation.
(2) Solid mechanics modeling, e.g., for jet engine turbine blades -- lots of math.
(3) Molecular spectroscopy: Some group theory and group representations can write down the possible spectral lines. That's long been the key to identifying molecules.
(4) My post mentioned Maxwell's equations for the design of the WWII radar that helped England win the Battle of Britain.
(5) A good, general purpose curve fitting technique is least squares spline fitting -- there's some math in there.
(6) Which pharmaceutical salesman goes to which physicians and leaves what samples to maximize revenue? Can be formulated as min-cost, capacitated network flows via the network simplex algorithm with a modification of W. Cunningham.
(7) Given a special scenario of global nuclear war limited to sea, estimate how long the US SSBN fleet would last.
(8) Given current revenue and planned capacity, project revenue growth -- solution saved FedEx.
(9) Given some ocean wave data, find the power spectrum, with confidence intervals, and generate sample paths with that power spectrum.
(10) At FedEx, which airplane goes to what cities in what order to carry all the freight, meet engineering, FAA, and safety constraints, arrive in specified time windows, and minimize costs.
(11) For an airplane, how to climb, cruise, and descend to arrive on time and minimize costs. Basically deterministic optimal control.
(12) Blue positions SSBNs and targets missiles in Red areas; Red attacks the Blue SSBNs; the remaining SSBNs fire. How best to position the SSBNs and target the missiles?
(13) A missile is chasing an airplane. What flight path should the airplane follow to evade the missile and what flight path should the missile follow to hit the airplane? A problem in differential games.
(14) A resource allocation problem in marketing resulted in a 0-1 integer linear program with 600,000 variables and 40,000 constraints. How to get a feasible solution optimal or nearly so? Solution involved non-linear duality theory and yielded a feasible solution bounding showed was within 0.025% of optimality in 900 seconds on a 90 MHz computer.
(15) My current startup.