I saw a blog post this weekend that looked at the IPO filings of 79 tech companies and calculated the ownerships of the founders and the VCs at IPO.
The result of that analysis is that the average founder ownership at IPO was 17% and the average VC ownership at IPO was 56%.
I’ve written a bunch on this topic and here are two posts that address this exact issue:
Founder Dilution – How Much Is “Normal”?
In both posts, I lay out how the equity gets shared with employees and investors as the company grows and scales.
Here’s the most important quote from those two posts:
In my experience, it will generally take three to four rounds of equity capital to finance the business and 20-25% of the company to recruit and retain a management team. That will typically leave the founder/founder team with 10-20% of the business when it’s all said and done. The equity split at 20% for the founders will typically be; 20-25% for the management team, 20% for the founders, and 55-60% for the investors (angel all the way to late stage VC).
I wrote that seven and half years ago, but on this topic, not much has changed over the thirty years I’ve been doing VC.
Raising round after round of venture capital is expensive. There are some entrepreneurs who figure out how to get profitable and not raise round after round (or avoid VC altogether), there are some entrepreneurs who are able to raise a very high valuations and avoid a lot of dilution, and there are many entrepreneurs who choose to sell the business before they take a lot of dilution. But for the entrepreneurs who raise four to six rounds of VC before going public, the math is the math. If you end up owning more than 20% at IPO, you are beating the averages.
Right. My first reaction on seeing the post on Techmeme was, “what’s new here?”I suppose every generation needs reminding.
Claim: Ambrose Bierce is Wrong. Wildly wrong. Uninformed. Ignorant.Proof: Go to a good research library. There go to the QA section. There find thousands of papers in peer-reviewed journals. According to the usual criteria, each such paper was judged by experts in the field of the paper as “new, correct, and significant”. So, there are lots of new things in the QA section. Done.Any questions?Example 1: A. Wiles settled the question of Fermat’s last theorem — for positive integers a, b, c, n, the equationa^n + b^n = c^ncan hold only for n = 1, 2. People struggled with this result since about 1670.Example 2: From the work at the LHC, the Higgs boson was shown to exist.Example 3: Penicillin. Saved lots of lives.Example 4: RSA public key cryptography. Cracking it is equivalent to factoring a positive whole number into a product of primes, and for large positive whole numbers we do not know an efficient way to do that. Crucial to much of digital communications.Example 5: For positive integer n, can calculate discrete Fourier transforms of time series of length n in O( n * ln(n) ) arithmetic operations. Due to J. Tukey with several important variations due to others. Revolutionized acoustic, sonar, and seismic processing.Example 6: For positive integer n, the fastest possible sorting of n keys by comparing keys two at a time is O( n * ln(n) ) (A. Gleason), and heap sort achieves this bound so, in this sense, is the fastest possible such sort. Unlike merge sort, heap sort is in-place.Example 7: For a positive integer n, the set of real numbers R, R^n with the usual topology, and a subset C of R^n closed in that topology, there exists a function f: R^n –> R that is 0 on C, strictly positive otherwise, and infinitely differentiable. Interesting cases of C are Cantor sets, Cantor sets of positive measure, sample paths of Brownian motion, and the Mandelbrot set. This result helped show that, in the Kuhn-Tucker conditions, the Zangwill and Kuhn-Tucker constraint qualifications are independent and settle a question in the famous paper in mathematical economics by Arrow, Hurwicz, and Uzawa. This result was published in JOTA in 1994.Example 8: L^2 is complete. An astounding result.Example 9: Every stochastic arrival process with stationary, independent increments is a Poisson process. Quite useful in practice.Example 10: Every L^1 bounded martingale converges to a random variable. Beyond belief, but it’s true.Example 11: The Plancherel theorem in Fourier theory shows the Heisenberg uncertainty principle in quantum mechanics.Example 12: With the scanning, tunneling electron microscope can image individual atoms.Example 13: In a Hilbert space, every closed set has a unique element of minimum norm.Example 14: There is a polynomial algorithm for the min cost capacitated network flow problem (D. Bertsekas). Used in US national security.Example 15: For the Gaussian distribution, sample mean and sample variance are sufficient statistics. Astounding result.Example 16: Random walk is recurrent in dimensions 1 and 2 but not in dimensions 3 and above.Example 17: Integer linear programming is in NP-complete.Example 18: There are no countably infinite sigma algebras.Example 19: Microelectronic lithography at 14 nm line width.Example 20: Microelectronic processors with 8 cores, 64 bit addressing, and a 4.0 GHz clock.One more:Progress in materials, structures, aerodynamics, and turbo-ram jet engines to fly at 80,000+ feet and Mach 3.1+ for 2000+ miles without refuelinghttp://iliketowastemytime.c…
“You didn’t invent sex”-JLM
.”Your generation didn’t invent sex.”JLMwww.themusingsofthebigredca…
Ahhh – but we found out about it – as most generations do.
.It seems essential to the passage of generations, no?JLMwww.themusingsofthebigredca…
Not well enough: We’re going extinct. Literally. But, think of the upside! Feminists have found things much, much, much more important to do than motherhood. Alas, such feminism can’t last!
Reminding of what? My generation didn’t even have a chance to make a bad deal.
First…generation doesn’t mean parent-child generation. But rather every cycle or class of entrepreneurs to go thru some incubator.Second…reminded that having to trade a large chunk of your company for cash to keep it running/growing isn’t a pay day.
To be honest, I never understood the obsession over percentages and dillution.Founders better maximize their ability to create value and scale it. Maximizing valuation might work adversely on the core purpose.
It’s interesting to note that with the new cryptocurrency offering models, the ratios are totally inverted. It starts with keeping about 15% roughly to founders and giving the rest to the decentralized community of stakeholders and users. VC’s aren’t part of the equation….unless they manage to squeeze themselves somehow.
giving? Au contraire mon ami. We ‘donate’ with hard earned btc or fiat, and prayers.You think 15% is the mark? Higher than I thought.Tendermint has talked about a giving model. I’m not sure if that’s still their preference.
So you are an expert in a segment and investing in one where you don’t benefit–I doubt it my friend.
Don’t get me wrong. Investors can benefit in Token models if they know what they are doing 😉
I wonder how these numbers are skewed for 2nd/3rd time founders. It strikes me that a more seasoned founder will not give-up as much to VCs over the course of the funding cycle dances.
More seasoned founders don’t give up as much to would-be co-founders either, especially if the would-be co-founder doesn’t pull their weight and EXECUTE on the list of their responsibilities and build to-do’s.As for VCs, well … they too have to prove they’ve earned the opportunity to invest by showing they’ve EXECUTED on the list of their responsibilities and value-adding to-do’s in their investments’ best interests.When everything aligns, it makes for easier ROI for everyone.
True but depends on the market success.How much has FourSquare given away? He’s an astute experienced founder with challenges as the market has moved.At the end of the day, you need capital. Sometimes at any cost.
Of course if things don’t go your way, you have to sell more at lower valuations.
Is this investing or gambling?
That’s the nature of V Capital.
If you can do diligence, it’s not gambling.
Well they have more table stakes a bigger stack. Also VC’s will ask for less. Also as Arnold says if you have runaway success you can name your price. In the slog?? A bit different.
Right, and typically they can get higher valuations baked in, as a starting point.
Examples of dual-class structures enabling founders with more flexibility to deliver on vision-mission:AmazonJeffrey Bezos47% FacebookMark Zuckerburg57%GoogleSergey Brin, Larry Page51%GoProNicholas Woodman43%And kudos to the one female:Care.comSheila Marcelo7% https://uploads.disquscdn.c…
Bezos owns 17% of AMZN, not 47%. And he’s the poster child argument against dual class structures, because he’s doing amazingly well pushing an ambitious vision at Amazon without owning dual-class stock; he owns Common like everyone else.
A-ha, thanks. The Blossom Venture’s analysis has him at 47%. I mistakenly recalled Amazon is dual-structure from this 2013 article:* http://www.forbes.com/sites…
Not sure I like or understand the methodology in that blog post.” we poured through the S1 filings of 79 publicly traded tech companies.” How were those companies chosen? 79 is not a large sample set. Median and average give equal weight to the billions owned by Zuckerberg and the 2% owned by the head of 2U, whatever that is.They should have developed aditional metrics for % of market cap, but either way, the sample size is too small to draw any worthwhile conclusions.
Was speaking with an acquaintance about this. They had built a multibillion dollar company in their lifetime. They said, “founders underestimate the value of equity”.It also brings up another point. It might be smarter for the founder to build a company that can get acquired for $20-$80M than to try and build a $1B company. They might make more money in the smaller acquisition, with less brain damage and quicker exit.
Of course you may be right. Of course in the real world none of this is really controllable.This is almost never a choice.Companies trajectories don’t exist on whiteboards they exist in the sweat of everyday bumping into the dynamics of the market.Decision by action by decision by action and on and on and on.
That’s why a lot of successful founders become investors after their first exit.
because they are venture capitalists in disguise. real founders want to grow a real business with revenues and profits, which is not what the ‘exit’ founder has in mind.
I mean even as they go back to build their second and third businesses.
William makes an interesting point about appcoins.founders start with finalised dilution and absolute clarity on their ownership at exit.all stakeholders are fully aligned and vested. And pulling in same direction from day 1.
Right. Distribution happens in the front-end, instead of gradually. Then, of course, the startup needs to deliver.
How so and why do you think this happens?
i’m observing the internal stresses of one or two blockchain startups and i see factions forming and reforming from month to month, with forks, spin-off ICO offerings, and general turbulence as people position and then reposition themselves for what they think will be ever quicker returns. it’s the primordial soup stage of block tech innovation, with human nature and instincts driving it forward.technically it is also the gamification of delegate voting systems. people think they can spot an opportunity to ‘grab’ value…and try to take it.Also, the block explorer is transparent, but it isn’t the legal record of coin/ token ownership. all kinds of interesting ownership transfer deals could be taking place behind the scenes, OTC, and UTC (under the counter), and entirely off market that influence alignment of interests.
Thanks, Jason. How do you think the governance structures can be improved to minimize gamification of the delegate voting system?
by observing where startups are getting it wrong, and then tweaking and refining the model of the technical architecture and the governance structure. it also very much depends on the nitty gritty of a project, no two being quite the same. this may be a space where going in early and deep is not necessarily the best move. Google was late to search but cleaned up. i don’t see the need to rush in. it’s all experimentation right now, and some of it will blow up.
What difference does it make?https://uploads.disquscdn.c…
Interesting. So with Square Dorsey has 24% but with twitter it says 2% (and has Costellos name) . Are these numbers correct? Maybe this explains why he is splitting his duties between two companies.
Jack was a minority co-founder at Twitter and the sole founder at Square
So all of small != a little of a lotIt seems this is about Numerators and Denominators and probability100 / 100 of small * certainty vs20 / 100 of big * riskIf we assume big / small = 50 – a goodish return is where say 20MM becomes unicorn 1 BThis implies for the founder (unpaid) risk of failure on road to unicorn needs to be less than 10%I reckon founder risk is MUCH higher than this – especially if you can reach break-even before the decision – as going for growth is risky.However – a founder may take some decent salary – I expect that is a large part of the untold story !Disclosure : Bootstrapped & Biased ! Totally unsure of what we do at 20MM value and breakeven.Doesn’t matter – it will be a nice problem to have !
thanks fred. helpful data
Throwing Rocks at the Google Bus, by Douglas Rushkoff.Chapter Four, Investing Without Exiting:- page 184, Investment Gamified: The Startup.- page 196, Ventureless Capital: The Patience of Crowds.Well worth the read.
Thanks, added to reading list.
Here in the Netherlands there is also another way – government lends you money, usually based on matching. You have to repay, of course but don’t dilute.
Hard to comment on founder dilution w/ out a thorough analysis of pre/post IPO valuation and market cap. Equity is meaningless w/ out context.
Fred, would appreciate your perspective on incubators and the level of equity they are vacuuming off the table. Recently saw a quality deal that had come out of a “lab style incubator”, where the idea in part comes from the lab management. The lab management are not active in the business, yet they had 65% of the common, this left the founders in a deep minority position before the first external dollar was raised. The team was looking for seed stage funding, so the funding road was going to be long. We passed based on the structure alone…
That is a new definition of “founder”, and it’s ‘founder delusion’.adventure badventure madventure
Yup.Reason to accept equity funding:Can make some parts of the work easier. E.g., for Windows Server and SQL Server, getting installed, configured, getting the security holes fixed, getting the system management functions understood and often scripted, etc., a few days of some consultants who do that stuff full time could help.Reasons not to:Dilution.Waste time trying to keep the BoD happy, the BoD that doesn’t understand the business.Face the fact that the responsibilities and loyalties of the VCs on the BoD are to their LPs, not to the company.Face the fact that if the business is good, then it is better than the average of what the VCs invest in so that VCs involved will lower the chances of success.Entrepreneurs might be interested in a really exceptionally good business, and so should the VCs. Alas, a good entrepreneur can plan such a success, but apparently the VCs, or their LPs, believe that such success is just luck so ignore all planning. So, the VCs would never take any planning seriously, and that could sink the business.
Diluted anything is so boring. I would rather work with a strong team with enough authority, control and energy to move forward than a weak team struggling for survival.My wish for the US today is that the winner of the election wins with more than 50% of the popular vote, undiluted and undisputed.
I agree with you so much about this.When authority is wishy-washy and the team has no clear mandate, that makes for half-hearted missions, plans that can’t be executed and potential that can’t be realized.
Great post. Interesting to see this in context of taking money off the table too. If your investors help you cash a little out earlier as per snapchat and facebook. Then you go on to push for IPO ignoring otherwise tempting offers along the way. Without this then maths might suggest an earlier trade sale is the way to go.
If you’re a founder, you only need to know this. Avoid venture capital at all costs. Its by far and away the most expensive money there is.