Startups require luck.
All successful founders (those that have built “Unicorns”) had great skills and great luck.
Skills need no definition, as they are widely understood and recognized. We know them when we see them, be it in engineering, product development, sales, marketing, or operations.
But luck is hard to define, so I use Michael Mauboussin’s definition. According to him, luck has two key attributes:
- It affects the startup or the founder and has a critical impact on their success or failure, and
- It is quite reasonable to expect that a different outcome could have occurred, if history were to be repeated again.
(for further explanation, see his fascinating and illuminating talk).
Startups have become more luck depenedent
It is well known that startup returns follow a power law distribution. However, where success falls on Skill/Luck continuum is not discussed enough.
This is what is happening to tech startups as of late. The shape of the graph alludes to power law distribution of startup outcomes.
Here’s where startups were, on the skill/luck continuuum.
Luck is playing a bigger and bigger role in startup success and this comes from the below 5 forces (apologies to Michael Porter) shaping the startup ecosystem.
Force 1: Clustering of Quality & Talent
The cost of building products has collapsed over the last two decades and will continue to collapse. Basic knowledge of how to build great products, such as design, product development, growth hacking and other previously “secret knowledge” has become widely (and mostly freely) available for those that are motivated.
The time and resource cost of building great products is lower than ever, driven by cloud computing, Moore’s law, and a liquification of markets for talent, all of which reduced upfront fixed costs.
The talent distribution between average and top companies is much narrower than ever, despite the frantic pronunciations to the contrary by the startup community about “10x talent” being a secret weapon.
Products have a more uniform quality now compared to the past. The quantum leap that the iPhone represented compared to the status quo was a rare phenomenon, and will get rarer still.
When was the last time you saw two startups with competing products that had an order of magnitude difference in performance or functionality? Slack vs Hipchat? Uber vs Lyft? The ease with which competitors can match a startup’s performance and function means that “talent” is not as much of a difference maker for startups as is commonly believed.
What does this tight clustering of quality at the top mean? It means that paradoxically, luck becomes more important than ever in determining the breakout winners.
Because most tech startups leverage non-linear properties like network effects, switching costs, proprietary data, and lock in to compound their competitive advantage over time, path dependence(and luck) will play a bigger and bigger role in success than objective parameters such as product quality, performance or reliability.
Force 2: Crowded Distribution Channels
Between 2008 and 2018, the number of smart phone users 10xed. Which meant that there were a lot of new users coming online that could be cheaply acquired. We saw a wealth of new products being built that leveraged these new user bases to create dense networks and killer businesses on the cheap. Think Instagram, Twitter, Uber and others.
But now, user acquisition is increasingly hard. ADWORDS has been unviable for most startups even back in 2012/2013, but now the distribution surplus Facebook/Instagram paid advertising yielded has essentially been exhausted, and CPMs have risen rapidly. This meant that existing companies with revenues or funding to compete for user acquisition can outspend new entrants consistently, making it difficult for startups to catch fire. This is borne in numbers.
From 2013 to 2018, the ratio of new to established startups at the top of google play (a surrogate marker for churn at the top) went down from 30% to 5%, as Eric Feng lays out.
This is to quite predictable and unsurprising. Facebook and Google are in a business of maximizing revenue per Ad, and when people get extraordinary return on their ad spend, the second place auction model of Google and similar algos of Facebook automatically drive the cost of Ads to their marginal value. While Insta/FB were new, there was still a lot of value to be discovered by accurately and creatively targeting user niches where competition was scarce, but with increased adoption and competition, the marginal costs caught up with marginal value quickly. This is a repeat of what happened with Adwords and the surplus provided by Facebook’s rich targeting has been mined quickly to exhaustion.
When distribution channels get saturated, it is harder for startups to breakout with out much distribution luck. Which again means that you need to get lucky to get to a critical mass of adoption fast. Either by finding niches outside of traditional distribution channels, PR, celebrity endorsements or other such factors.
Force 3: Incumbents are better at encroaching
This should be the most worrying trend for startups. FAANG and other incumberts can now quickly create imitation products and use their scale and distribution advantages to nullify first to market advantages of startups.
Incumbents are now increasingly better at cannibalizing new, adjacent markets. See how Instagram stories lead to a quick nullification of snap’s competitive advantage. Snap stock never approached it’s IPO price. Meanwhile, Facebook’s almost doubled. From a base that was more than 10 times larger!
Contrast this with Facebook’s reaction (or the lack of it) to Twitter’s rise. Or compare how Microsoft tried for years to catch up with Google in search, did technically, but failed to replicate the ecosystem effects of adwords SEM industry. But as soon as it perceived a viable messaging platform in Slack, it leveraged Office 365 integration and footprint to make Teams a viable competitor in short order.
Even giants such as Netflix aren’t safe anymore, as Amazon and Apple have leveraged their user base advantages to effectively compete with the category creator and market leader.
Competition between Booking.com and Airbnb in vacation rental market is another example where conglomerates can leverage their scale and inventory advantages to cross sell and encroach on the innovators’ turf.
There is a historical precedent for this kind of dynamic. The closest thing to take a look at is the evolution of another hits driven business. During the studio model of Hollywood which started in 1927, (until it ended in 1948 with the federal government antitrust legislation), a few studios came to dominate film making.
They did this by not only producing movies in their own studios with their own equipment and having creative talent on long term contracts, but also by dominating the distribution channels through vertical integration and pricing power. This is very similar to the current FAANG tactics.
Again, when incumbents coopt distribution channels, you need more luck to breakout and escape their frontal assault and get to critical mass fast.
Force 4: Capital for late stage, but not for early (“Softbank Phenomenon”)
An abundance of capital available to the top companies in a vertical means that the top one or two companies with the resources to out market their competitors and steal customers get an unassilable distribution advantage and lock in. Especially in winner take all/most markets, a lack of capital is deadly, when your competitor is willing to lose a lot of money to steal your potential/current customers.
This is shown by the recent trend of an increase in the size of growth rounds disproportionately, compared to Seed and A/B.
Seed rounds have increased in absolute numbers, but the conversion from seed to series A is way down because series A numbers haven’t kept up. From Seed to series A, to B and later, with every round, the odds of successful exit increase an order of magnitude or more. Given this fact, most funds opt to increase their AUM and move into later and later rounds (also to increase their management fees) and increase their odds of returning 3x.
Ultimately this means that capital will flee seed stage companies as the odds of portfolio companies getting over series A hurdle continue to go lower and lower.
This brings us back to the same question: How do you get lucky and avoid the Seed to Series A death trap?
Force 5: Access to late stage liquidity to early investors
Not only are seed stage startups competing with other seed stage startups for capital, but with late stages companies whose odds of success are much better than their own. Traditionally, the only way for non institutional investors such as HNWs to get exposure to startups is to invest in seed stage startups and take soak up the massive amount of risk. But, with Angellist syndicates and secondary markets such as Equityzen and others, they have better options.
If you are an angel or a micro VC that can deploy $1M in 2 years, you can now pick between
a. funding 40 seed stage deals, or
b. buying equity in 40 Pre IPO companies using secondary markets, or
c. investing in later stage rounds of post seed companies via angellist syndicates.
Given that the success odds are much higher in later stage deals, The natural choice would be to do b or c, if your goal is to reliably 2–3X your fund. This will mean seed stage capital will migrate out.
Startups resemble Movies and Music
All of the above 5 forces acting together means that startups will increasingly resemble hits driven businesses like movies and music.
If you are a founder thinking of starting a tech company, your risk calculus is quite different now compared to 10 or even 5 years ago. Luck will play a larger role in your life as a founder.
Given this, would you follow the conventional silicon valley wisdom of finding a problem you are passionate about solving, and spending 7 years to make it work, knowing that luck will continue to be the critical missing variable?
Or, is there a better way?
(Read part two to learn how to engineer your luck as a founder)