Skip to main content

It’s not where you take things from — it’s where you take them to.

Jean-Luc Godard

Just over two decades ago I remember I had my first real product idea. It was going to be a “SMS over GPRS” idea named “GSMS” to save me money on high SMS charges. I built the first version on J2ME (A Java framework popular in the early 2000s) and I remember asking people sign NDAs in order for me to reveal the world changing idea I had.

A clear sign of an amateur is over valuing their idea. There are exceptions in screenwriting, BioTech and the Material sciences but these are rare and very specific to their fields. The central flaw in this overvaluation is the thought process that ideas have a common endpoint, that if they are copied they will be copied in the same way and that most of the value lies in the concept and not the execution.

The reality is that Ideas take a much more divergent path. A company’s state at any given point in time is a sum of many ideas, pivots and actions over time, and this is why two companies serving the same market end up performing and looking very different to each other. These actions are not taken so much as they are caused. Why a certain action is taken can somewhat be disassembled into:

  • What inputs are considered or looked at with frequency?
    • Are you looking at competitors as inputs, customer behavior, A/B tests or ignoring inputs and building towards a point in the horizon at the cost of everything else.
  • Biases based on the team’s past experiences, emotional state, and personal desires for the company both in the short and long term

In most cases you can’t steal ideas because ideas have to be created and there are many decisions to be made which make the final forms divergent.

These decisions over time become the company, sort of the same way that the decisions within your life define you. You are a sum of your parts as a team is a sum of their parts. Unfortunately the best decisions and prioritizations can fail if the market is wrong or the timing doesn’t work. They can also fail due to black swan events (high impact, low probability) such as theft, hacks, bad Venture Capital.

So building a company in some ways is acknowledging these effects that are difficult to predict and obtain a fine-tuned level of control over. Instead of trying to make sure you’re making the right decisions you could make sure that the INPUTS into decisions become more effective even when things are not going well. Or as Nassim Taleb would say, Anti-Fragile.

Who you hire is the company you get

If you hire people who match the shared values you’re trying to create this can be a resilient strategy. If those values are brittle (greed, prestige) then they will amplify any difficulty as compared to reduce it. Avoid people who want to be part of a “hot new startup” as their primary motivation. There is a overly simplistic but useful truth around “Learned needs theory” which states that all of us seek Achievement, Affiliation and Power in different quantities.

Achievement (nAch): This is the drive to excel, achieve in relation to a set of standards, and strive to succeed. They are often motivated to pursue difficult problems to improve their own abilities, and often preferring to work alone. They work well with high risk and uncertainty.

Affiliation (nAff): Individuals with a high need for affiliation value building strong bonds, being part of a group, and are concerned about maintaining harmonious relationships and can be demotivated with high risk and uncertainty and the need for prestige (in extreme cases)

Power (nPow): This is the need to make others behave in a way that they would not have behaved otherwise. People with a high need for power want to control and influence others. They seek positions of authority and are often more concerned with prestige and gaining influence over others.

I think for a startup you want to optimize for hiring people who are primarily motivated by Achievement but have affiliation with those they respect, for example have created strong bonds and relationships with their colleagues. You want to avoid those who entirely seek prestige and completely avoid those who seek power. Maybe the proportions you want are 80/20/0

Tight Customer Feedback Loops

Everyone says they are customer-centric but in reality finding time to always talk with customers is hard. It becomes easier to play house than build the actual house and prestige will always pull you in this direction.

I recently had two calls with customers, the first was excited about the company, what the product could do and the impact we are going to have on AI and the world.

The second was with a customer who thought I was a support agent and was frustrated with not being able to get the results they wanted.

My ego enjoyed the first call much more but it was useless, they didn’t even really buy the product to use it but rather to be part of something new. The second customer bought the product because they needed something done and they didn’t care who I was or how many other customers we had serviced, they wanted their job done. I’ve found that the most unhappy customers are my greatest opportunity to learn something new.

The more you drink the Kool-aid about how amazing your product is the less likely it is you’re going to improve it in meaningful ways. Tighter customer feedback loops means:

  • Talking to a variety of people are different stages of their journey (unaware, consideration, onboarding, usage, retention etc.)
  • Asking good questions
  • High frequency (doing this weekly)
  • Sharing it and consuming it internally

This is anti fragile because even if things are going bad you are getting higher levels of information as to why and corrective measures can be taken

Just as an executive who boards themself up in a corner office with a PA separates themselves from their company and leads to failure so moving yourself away from customers will lead to failure.

Metrics – You are what you eat

A deliberate choice of measures that represent the true value of your product to the customers. In the same way in that you are what you eat you are what you measure.

The blurry line comes in what is best for the business vs what is best for the user. It’s hard because you might be trying to move organic sharing or referral mechanics while decreasing value for the user.

It gets worse if you structure and incentivize a team on specific metrics where they are not being evaluated on addressing a user’s pain point or need

The measures do not need to be comprehensive or without error but they need to be closely linked to what the users value and have directionality with a sense of the levers that could be pulled to improve the measure. Selecting the right measures:

  • Outputs not Inputs. Although it’s helpful to measure inputs in areas such as sales (# of meetings per week) key metrics you want to look at should be linked to user’s behavior.
    • These outputs should then be filtered for those which are user value orientated not company value (e.g. “time spent on app” which is a company value metrics but not an indication of the quality of time spent). Sometimes low engagement can be high user value and high engagement can be low user value.

If you don’t yet have a clear measure of the success of your product with a user then you should work on building a model of it.

There are more anti-fragile principles I intend to explore at another point such as creating space for failure, red-teaming and rituals. Unfortunately there are no principles that hold true in all situations but what you can do is decide where you don’t want to fail.

“Don’t worry about people stealing your ideas. If your ideas are any good, you’ll have to shove them down people’s throats.”

Howard Aiken

Copyright 2022 - All rights reserved