Rationality, and Startups

How to avoid the most common fallacies in early stage startups?

  (The following is a rough overview of Spencer Greenberg's excellent presentation on Rationality in Business, with a couple of examples.)


  The motivation for applying rationality in businesses is straightforward: the stakes are high, and the incentives are aligned so that more rational (goal-seeking) decisions leads to better performance, and more success. In all honesty, most founders try their best to be rational; however, the time between decisions, and consequences are so long, that it won't show itself until it's too late.


  Frequency map on the failure modes of web Startups:

  • 1. Not building something people want
    Antidote: Talk to your customers, list, and validate your assumptions. Do rapid prototyping ("fake it"), and find ways to demonstrate demand before building it.
  • 2. The consultancy trap: having business models, that scales with the number of sales + engineer headcount
    Antidote: build renewable audiences, and optimize for repeatable sales processes.
  • 3. Unit economics: negative transaction, having a customer acquisition cost higher, than the lifetime value of customer; negative cashflow
    Antidote: build, maintain, and communicate articulately a rough cashflow model of the business, with real numbers.
  • 4. Team domain see below


  These are usually (but not exclusively) stemming from faulty thinking processes; specifically, and sorted by venture progress:


  1. Coming up with ideas

  • Irrationality: Not sitting down to seriously think it through.
    Most entrepreneurs tend to fall into the trap of crusaders: having struck with a light of inspiration, they proceed to execution immediately; which results in local optimization, as opposed to global maxima.
    But the value of spending 5 hours on thinking the idea through can not be overestimated. Most failure modes are easily avoidable via simple financial modelling, Fermi estimation on market size, using this estimate to set a target range on pricing and cost, and taking the outside view on current consumer behaviour.

  • Irrationality: Sunk cost fallacy
    Even after 2 weeks of working on the idea, sunken cost fallacy makes it mentally obnoxious to switch to a better idea.
    On one hand, pivots should be approached with caution, and made such that maximal number of assets, and *insights* carries over the new direction. On the other, updating on new evidence allows radical trajectory change.

  • Irrationality: "But ideas are worthless, execution is everything"
    Ideas are force-multipliers: a small change in the conception phase makes a huge difference during execution. Consider taking seriously that you are about to spend multiple years on the project, and plan accordingly.

  • Irrationality: Not bothering to make predictions
    While most forecasts are in fact pulled out of thin air, going through the motion of forecasting has a high VOI (Value of Information): it can fundamentally alter your target market, and product line.

    Specifically think through:
    -Do you have a competitive advantage? (Is it sustainable in light of rapid copycats?)
    -Do you understand the target market? (Can you tell the 3 phrases which makes you instantly seem "in-group" with them?)
    -Is there a large enough market? (Basic calculations, eg. targeting a total addressable market Fermi-estimated to be 1000, with a one-time $10 product might prove to be suboptimal choice. )


  2. Green-lighting it

  • Irrationality: Overconfidence: Most startup founders, when cornered with giving a probability estimate on *their* venture's future success tend to grossly overestimate. This is probably a selection bias -would-be founders with low probability estimates tend not to embark on the journey in the first place. The "outside view" on this is sobering:
  • 50% of all US businesses fail before year 4
  • PG: 90% of web startups fail, period
  • 80% of VC funded web startups fail (*after* having to go through a rigorous selection process)


  3. Selecting a business partner (or lack thereof)

  • Irrationality: Going alone. Not because it's impossible, but outside view estimates the process of reaching product-market fit to be 3.6 times slower in single-founder businesses
  • Irrationality: failure to optimize co-founders #1: getting people with your general skillset on board, as opposed to optimizing for complementary skillset. Too many times you see 3 engineers working together, without any sales, or marketing plan (see Founder's Dilemmas for more on this trade-off)
  • Irrationality: failure to optimize co-founders #2: considering how people behave during "ordinary days" as a basis for the relationship. The Startup journey is anything, but ordinary: opportunities for screwing others over are abound, and the temptation is high. Consider the worst facet of your would-be co-founders while working on your roster.


  4. Disclosure

  • Irrationality: paranoia. Copycats will copy your site; patents offer little to no protection, esp overseas.
    On the other hand, idea disclosure can lead to many positive opportunities, either by seeing different perspectives on the matter, or lining up potential users.


  5. Negotiation

  • Irrationality: *Anchoring effect*. Be careful of naming a number: studies show successful deals to be roughly +-20% around the first number named. (Patrick McKenzie advises not to be the first in naming a number during consulting negotiations).
  • Irrationality: Failure to negotiate. Too many times founders aren't asking for a discount, or averse engaging in making the deal sweet for themselves, leaving money on the table. But social rules hardly apply for business engagements: negotiating takes 5 minutes, result is accumulating over several years.
  • Irrationality: Responding emotionally to negotiations. All too often founders fell pray to defensive triggers during the most critical parts of business negotiations
  • Irrationality: Assuming other people will reach rationally
  • Irrationality: Applying normal social rules to negotiations
  • Irrationality: supply-side positioning&pricing. Your offering's position, and price has to be derived from the benefit your customers are getting from the services -as opposed to "how much it costs + margin"-type calculations


  6. Delivering the product

  • Irrationality: Planning forecasting. "This should only take a day!", "I'm in a hurry, so I should be able to get there in 30 minutes". Do reference class forecasting! Ask how long similar class problems took last time.
  • Irrationality: Failure to check assumptions against reality. "We'll launch next month!" for the past 3 months. To avoid: launch smallest-subset features rapidly, and iterate -you won't know how (whether) customers behave until it's actually in their hands.


About the author: Joel runs a product inbucation, and consulting company in San Francisco. If you'd like to read more by him, subscribe to the product development mailing list.