Most startups fail. This is why the equity offered by startups often feels like a scam to applicants and why we created a better startup equity plan.

The Allure of Early-Stage Startups

We yearn to be founders, or at least employees at early-stage startups. We aspire to conceive something new, then implement and market a product the world has never seen. We should be encouraged to try. Ideally, the price of that attempt would be less discouraging than years sleeping on an air mattress eating ramen noodles for all three meals and the probability of success would be higher. The world is not always ideal.

The problem is that most startups fail.¹ A variety of sources peg the failure rate around 90%.² Worse, in the 10% of cases in which a meaningful exit occurs and the employee sticks around long enough for their shares to vest, the payout is orders of magnitude more likely to be hundreds of thousands of dollars, rather than millions, and orders of magnitude more likely to be millions, rather than tens of millions. In other words most of the “successful exits” (the 10%) do not yield enough cash to offset the lesser salaries, let alone the fantasized seven or eight-figure payouts. Unicorns are almost as rare as the creature for which they are named.

This is not the type of tail one should chase.

Same Upside, Better Odds

We created something new at AE and we suspect our model offers a superior alternative.

Firstly, our clients provide an extraordinary window into the tech landscape, and their needs often shine spotlights on fertile ground within the marketplace. For instance, when a potential client requested a custom platform for hosting health and wellness video content, we offered to develop it ourselves (much cheaper for them) with the client as our first customer. They prepaid for the first year of service, validating our assumption about the product’s value.³ As our first customer, they could ensure that the software met their specific needs. Developing with the assurance of a paying customer de-risked the venture dramatically.

A few months later, customers are pouring in, with $11K in monthly recurring revenue, up several thousand from the previous month and continuing to grow. Instillvideo has served over 20K users and processed millions of dollars since its public launch.

Secondly, conflicts among founders represent a common cause of startup failure. Forming a startup with colleagues already known to be amenable collaborators offers an additional sense of security and a higher probability of success. We teach the art of founding, build reliable external structures around our people, and use the surprisingly insightful PDP tools to meld personality types into successful teams. And in a (growing) company of 130+ laden with creative, hungry (startup-ey) folks all refining their skills through client work (many of whom are startups themselves), the insight, wisdom, and resources for success are all available. Who is more likely to succeed, all other things equal, the founders immersed in the expertise of what to do when or those going it alone?

Thirdly, AE builds long-term thinking into the DNA of every product it develops. This often means that when we explore an idea and de-risk with the sale of a potential first customer, we’re building and selling a startup with considerably better odds of a lucrative exit. For instance, we developed software to allow users to manage subscriptions with a text message. In the short-term, customer value decreased as some users elected to skip a month or lower their monthly purchase. In the long-term, those customers become more durable, more valuable long-term partners. Taking the long view is good for human agency, and great for business. We recently sold ElectricSMS to Recharge for $6M and a pile of equity.⁴

The back of the envelope

So, would you rather be a founder of an AE Skunkworks project or the 1st engineer at a startup?

Let’s define some parameters, build a simple model, then (gasp), do a little algebra.

Let’s assume the size of exits are exponentially distributed⁶, e.g. a probability density function of the form:

Thus, the cumulative density functions⁷ are given as:

And the expected value of any of these distributions becomes:

Therefore, the expected value of an AE Skunkworks project is given by:

And the expected value of an traditional startup’s equity is given by:

Of course, we do not know these values precisely, but we have some intuition. The probability of a startup exiting is ~10%, as a rule of thumb. The proportion of equity given to an early-stage startup engineer, after dilution, could generously be estimated at 1%. AE offers, say, 20% to its Skunkworks founders.⁸

Let’s start talking about break-even points:

Now let’s fill in some of the more easily estimated values:

Next, let’s cleanup a little:

What does this mean?

If you believe that the expected exit size of a de-risked, first-customer-pre-sold, AE Skunkworks project is comparable to the expected exit size of any traditional startup, then you believe those lambda terms cancel, which leaves:

In other words, if you believe that the probability that a AE Skunkworks startup ultimately leads to an exit is better than 1 in 200, AE offers you a more desirable expected outcome. And since our track record suggests that P(Exit | Serious) is probably at least an even proposition (a majority of startups fail due misreading of market demand or a poor group of founders, both of which AE’s model largely eliminates)⁹, you’d really only need to believe that the probability of a Skunkworks idea becoming “serious” is at least 1% for AE’s model to be preferable.

But wait, there’s more!

Try, Try Again

At AE, one earns a competitive salary while developing their nascent venture. At a startup, one accepts a below-market rate in exchange for the equity they are receiving and the promise of future riches. But that’s hardly the most important distinction.

At AE, some Skunkworks projects never locate that first customer or otherwise do not demonstrate the promise required for their continuation. In other words, the losing side of the pserious binary outcome.

In that case…try again! In the traditional startup paradigm, one basically is offered a single trial of the expected value calculation above, for which one pays with 4-5 years of the most energetic professional time of their lives. The alternative is to roll the dice multiple times until p_serious lands in your favor. At that point, the break-even math is simply:

There is no question that the probability of an exit, for a startup incubated amongst long-term thinking, talented, amenable colleagues, enhanced by the perspective of diverse client work, and in possession of a pre-sold first customer is well in excess of 1 in 200.

Most startups lack the majority of those features, and the probability is still roughly 1 in 10.

Oh, and we also receive equity from other client work and incubated Skunkworks projects whilst you try, err, learn, grow, and retry. In other words, you succeed if any of the serious Skunkworks exit, not just your own.¹⁰

So if you find yourself pondering whether becoming an engineer at an early-startup or an engineer in an incubation environment like ours is preferable, we ask you kindly to “do the math.”

We suspect you have more questions about AE's equity plan, and how this might compare to alternative offerings of traditional startups. For more entertaining equations, wordplay, and witty discussion, we encourage you to read this essay!

¹

^{Pretty sure I already said that, but it bears repeating.}

²

^{Here’s one such publication of that statistic from Investopedia, but the 90/10 rule shows up frequently.}

³

^{We also researched the market, confirming that there were numerous similar customers who would pay for such a product if it existed!}

⁴

^{We are legally unable to disclose the exact dimensions of said pile. We really wish we could, and we think legal systems as they exist today should be innovated upon a bunch sometime, but it’s not on our immediate roadmap.}

⁵

^{For the sake of argument, let’s assume these parameters include all necessary dilution (which does mean that the traditional startup that offers you 2% equity at some nascent stage, by the time the exit occurs, will offer you far less, probably more like 0.5%)}

⁶

^{Yes, the distributions of startup exits probably have heavier tails, but exponentials are far easier to work with, have well-specified expected values (convenient for backs-of-envelopes), and possess the needed property. Specifically, the probability of a given size outcome diminishes exponentially as the size increases.}

⁷

^{These are included in case the preferred heuristic is the probability of ultimately receiving a $1M (or other) payout rather than maximizing expected value. Though I fear creating some post-traumatic stress in the minds of the ex-engineering academic audience, I leave that analogous procedure as “an exercise for the reader.”}

⁸

^{The actual value could be higher or lower, depending on the number of founders involved.}

⁹

^{Between ElectricSMS’s sale, Instillvideo’s explosive growth, and some of our nascent ventures, the empirical ratio is defensible. To hammer the point home, when culling the internet for startup statistics, one finds that 42% of startups misread market demand and 23% fail due to a weak group of founders. The former is impossible if we’ve already pre-sold a paying customer and research reveals numerous others. The latter is improbable after experience working together and PDP’s blessing!}

¹⁰

^{Full disclosure, like other startups, if an employee leaves, AE reclaims some of their equity to ensure that new employees can partake. Unlike other startups, if the employee departs for one of the many companies in which AE owns equity (including one they themselves founded whilst at AE), they keep all the goodies. Moreover, if an internal startup is acquired, and the employee departs to work for the acquirer, they also keep the goodies. Extra incentive to found a startup of your own while you’re here!}

No one works with an agency just because they have a clever blog. To work with my colleagues, who spend their days developing software that turns your MVP into an IPO, rather than writing blog posts, click here (Then you can spend your time reading our content from your yacht / pied-a-terre). If you can’t afford to build an app, you can always learn how to succeed in tech by reading other essays.