Estimating Startup Complexity by Assumption Count

Reading Time: 2 minutes

Questions addressed in this post:

  • How do I choose between startup alternatives?
  • How long before startups reach profitability?
  • Which startup idea is best?

All of these questions require a second question: “how complex is each startup idea?”

In “Lean Startup“, Eric Ries defines “runway” as the number of pivots a startup has remaining before it runs out of cash. This implies you could calculate a startup’s funding requirements by estimating the number of required pivots (provided the cost per pivot can be estimated, possibly from burn rate and frequency of “pivot or persevere” meetings).

A startup would only need to pivot when it realizes one of its key assumptions is wrong; for example, if the value hypothesis is invalid and nobody wants to pay. This is realized when the actual metrics don’t reach the target metrics, despite optimization attempts.

Ideally, a startup would pivot along until it finds the nearest viable model (which may no longer be an innovative one), but as Ries suggested, every pivot might be a startup’s last as it eats runway.

Since these costly pivots are caused by assumptions, it follows that the risk, complexity, and cost of a startup must be largely determined by the number of founding assumptions:

First-order startup complexity  ∝   # of founding assumptions

A zero-assumption startup would be purely replicative: selling the same product to the same market under the same conditions as an existing successful business. Each additional assumption — such as those related to problem, product value, growth, or competition — creates a potential pivot, which increases expected startup risk, complexity, and cost.

Since every key assumption must be correct to successfully implement the founding vision, the probability of success is proportional to the power of the number assumptions:

Probability of success  ∝  c(# of key assumptions)

Effect of Key Assumptions on Probability of Startup Success

The chart illustrates a simple example. The chance of a startup eventually succeeding is higher than shown because it can pivot around a wrong assumption.

Since the reality of business requires selecting from among competing opportunities, carefully counting founding assumptions may be a useful evaluation tool.

Get new articles by email
(happens rarely!)
You will only get new posts.

How NOT to launch a new product: A $4 Billion lesson in 2 minutes

Reading Time: 2 minutes

When is a product launch a “sure thing”? How about when you’re the market leader, employ an army of marketers and designers, enchant the media, and invest four billion dollars?

Ford had every advantage, and the result was disaster. What went so wrong?

It was the mid-1950s, and Ford needed a new medium-priced model to round-out their line. They invested heavily in marketing research to determine what consumers wanted in a new car. They meticulously analyzed the competitive landscape. They designed genuine innovations that would later be copied by other manufacturers. In short, they did everything one would think sensible and prudent.

Yet the launch was a complete flop. Manufacturing was halted within three years — long before reaching break-even sales — and the “Edsel” name became synonymous with monumental failure.

If they did such excellent planning, where did it go wrong? How can we avoid making similar mistakes with our own product launches?

1. Ford paid lip-service to their customer’s preferences, but disregarded them on key points. For example, they produced an incredible list of 6,000 name ideas ranked by consumer reactions and expert opinion, but top executives couldn’t pick a winner. The puzzling “Edsel” brand (a low-ranked name in their research) was chosen by the chairman of the board to break the debate. As John Brooks’s concluded in his 1969 essay on the topic,

“Although the Edsel was supposed to be advertised, and otherwise promoted, strictly on the basis of preferences expressed in polls, some old-fashioned snake-oil selling methods, intuitive rather than scientific, crept in.”

2. It’s hard to believe that in a $3 billion development budget ($250 million in mid-50’s), basic test marketing was never conducted. This can be done from a very early stage in the form of concept testing: showing a finished concept to the target market and getting feedback. Done quantitatively, several concept variations can be compared to determine the best one. In a later stage, consumers can try prototypes. Ford took the opposite approach: keeping their Edsels literally “under wraps” and hidden from public view until the grand unveiling.

Edsel Carrier

3. Watch people shop for as much as a toaster, and it becomes obvious that pricing is crucial. The Edsel’s pricing was confusing, overlapping with other product lines so consumers didn’t know where it was supposed to stand. Furthermore, the positioning was unclear: highlighting innovative features that could only be considered “premium”, while lacking features common to other premium cars.

4. For over a year the Edsel was slowly revealed to the public in what some called and “automotive striptease”. It was hyped through stunts and ads like no other car had been, so expectations were naturally high. But after release, the only sensible media story remaining was “does it live up to the hype?” With such an intense spotlight, minor problems typical of any new model became headlines, fueling a devastating negative news cycle.

Key takeaways:

  1. Apply your marketing research at each stage of development and marketing.
  2. Test market early and often to avoid launching an “Edsel”.
  3. The consumers’ purchasing decisions consider the competition, so you should too: position clearly.
  4. Don’t build unrealistic expectations or over-hype your product: it invites piercing scrutiny.