Premature Scaling Kills Startups: the Genome Data

The new Startup Genome project tries to bring structure and data to the following question: why do so many companies die from premature scaling ? I was impressed by the first report the Startup Genome team produced so when Bjoern asked me to comment I was more than happy to comply.

The first Startup Genome Report dealt with understanding causes of failure during the Lean, iterative phase of a startup's lifecycle.  This time, they decided to attack the complex and more nebulous area in a startup's life when a company decides that it's ready to scale and gets it wrong.  All companies get it wrong to a certain degree and as Ben Horowitz rightly pointed out, "product-market fit is not a discrete event".  [Reminder: If you only read one blog, read his].  The interest of the report lies in looking at data and trying to systematically determine when these errors lead to ultimate failure.

In their own words:

Startups are temporary organizations designed to scale into large companies. Early stage startups are designed to search for product/market fit under conditions of extreme uncertainty. Late stage startups are designed to search for a repeatable and scalable business model and then scale into large companies designed to execute under conditions of high certainty.

A startup can maximize its speed of progress by keeping the 5 core dimensions of a startup Customer, Product, Team, Business Model and Financials in balance. The art of high growth entrepreneurship is to master the chaos of getting each of these 5 dimensions to move in time and concert with one another.

Most startup failures can be explained by one or more of these dimensions falling out of tune with the others.In our dataset we found that 70% of startups scaled prematurely along some dimension. While this number seemed high, this may go a long way towards explaining the 90% failure rate of startups.

I encourage entrepreneurs to think through the full Lifecycle all the time and I think this is a great initiative.  I jotted down some of the reasons why I see companies scaling before they're ready to throw some quotes at Bjorn:

  • Illusions of product market fit or price discovery: The classic mistake is to confuse a few early adopters with a market.  This was perfectly documented by Geoffrey Moore and is still relevant today.  Not being able to replicate early successes also occurs in consumer businesses, usually through the rapid degradation of customer acquisition and conversion economics.  It's one thing to spend $50K a month through online marketing channels with high marketing efficiency, but quite another to spend 5 or 10 times that amount without shooting the economic contribution of each new user or consumer.
  • Confusing Founder Heroics with a Repeatable Model: As Peter Lynch famously said, "go for businesses that any idiot could run, because some day one will".  The parallel here is that it is an easy mistake to assume that hired salespeople or marketing people will be able to replicate the same rate of success as an incredibly motivated, tenacious and compelling founding team did in the early days.  Of course that's rarely true.  The commercial proposition needs to be baked enough and repeatable enough that possibly less talented and less determined sales organisations will be able to shift product efficiently enough.
  • Unprofitable Scaling / Absence of operating leverage: You may be able to sell more faster, but you might not be ready to scale profitably.  The more customers you get, the more bugs, new requests and support issues start to bog your infrastructure down and ultimately come back to rest on the door of your salesguy.  I have lived through one experience of a company that had not understood that it had a total quality problem on its hands and kept ploughing money into the front of the funnel, to no discernible end.
  • The Tail Wagging the Dog (Board Pressure): You've raised a fair bit of money on the back of an ambitious set of projections, anod now you feel compelled to hit these numbers.  Never mind that the market may change or that you may discover these new verticals aren't as easy to crack as you thought, it will be difficult to go back to the board and tell them you won't spend the new cash, or hit the numbers.


The Priebatsch versus Lauzon debate, as old as startups

Scvngr's Chief Ninja, speaking at the recent Nantucket conference, clearly made a few people uncomfortable with his (fun) pronouncements about how every startup should raise as much money as early as possible.  Whilst I enjoyed his wit and style, I was twisting in my seat, though nowhere near as much as the other young founder on the panel, Matt Lauzon.  He seemed a little sick. 

When Matt took him up on it and said he would never advise founders to do this, there was a perfect Green Goblin versus Spiderman moment between Mr Orange who's clearly going for a $1BN or bust, where the management team is made up of a Chief Ninja, a bunch of Rockstars, a Pixel Czar and so on (did I mention the King of Bling) whilst Mr Black patiently incubated at Highland and puts his Customer Support Gals front and center of his website.

The point on these two is that you have, on the one hand, a momentum player who wants to own the market yesterday and will throw everything he's got at it, and another who's on a consistent and possibly more mundane path of building team, culture, repeatability (and talks a lot less about it).

Style aside (I do like a bit of camp in my startups), I know where I'd put my money and I'd certainly never advise young founders to raise as much as they can as early as they can, but let's just look at what the data says.

Report Conclusions

You should really get the report, but let's just say that it reads as a homily to

(a) being highly consistent, measured and systematic in progress along all the core axes of execution (customer, product, team, finance, mode)

(b) constraining yourself until you're really hurting for resources, including people and capital

Borrowing from The Master (in Germany: Das Guru) on the excellent RWW

  • The team size of startups that scale prematurely is 3 times bigger than the consistent startups at the same stage
  • 74% of high growth Internet startups fail due to premature scaling
  • Startups that scale properly grow about 20 times faster than startups that scale prematurely
  • 93% of startups that scale prematurely never break the $100k revenue per month threshold

Before the Lean Startup was codified so beautifully (the elegance of the theoretical model has to make you smile, does it not ?), we at Atlas used to call this Prove / Build / Scale.  It was not a perfect framework, and initiatives like this are key to bringing a rigorous and systematic approach to company building, so it can be taught as a form of management science.

Me ?  I will read, learn, and do as usual: forget all the lessons in the world and go with "data-augmented instinct" in my decision making.

Go crush it.  Let's make these stats better.

This entry was posted in Entrepreneurship, Management issues, Startups. Bookmark the permalink.

6 Responses to Premature Scaling Kills Startups: the Genome Data