Innovation Strategist & 2x Entrepreneur

An Innovator’s Logic

This is the first of a 3-part series discussing what I’ve learned are the three most important skills for innovators: Innovator Logic, Innovator Eyesight and Innovator Persuasion. Over the past few years, I’ve been neck-deep in the innovation world inventing products, services and experiences for the world’s most prominent brands. This is what it takes to create truly new things.

Creating ‘new’ things is inconceivably difficult.

Have you ever heard someone claim they have the next big breakthrough? They are going to launch to fanfare and make a billion dollars? All they need to do is raise venture capital or convince the CFO to fund the project, and the world will never be the same again.

Not so fast cowboy.

My question is this: have you found the edge of the body of human knowledge, and what have you seen or done to extend it?

Most of the time, that question is met with a this:

Here’s the bottom line: often we mistakenly believe we have created something new, only to find it’s a rehash of something that already exists.

True innovations can only occur at the edge of human knowledge, where an innovator has found the edge, extended it in some way, and then built a new product/service on the basis of that fresh perspective.

This post discusses an extremely powerful way to achieve that: first principles reasoning.

To state the obvious, Ideas are not Innovations.

Any group of people can sit around with whiteboard markers and sticky notes and come up with ideas, but when you boil it down, brainstorming is just an exercise of self-satisfaction. Most of the time, brainstorming is just group of people in a room trying to impress one another with something no-one in that room has ever heard about before.

But the context of human history extends beyond the significance of the 3 hours you spend with your whiteboard markers. It extends thousands of years, documented in untold volumes of books that represent human knowledge. For a group of people in one room to think they have come up with something completely and entirely new is simply delusional.

Innovations take a different form: they solve a problem that has never before been perceived, with a superior solution that has never before been conceived, or through a method that has never before been designed. Innovations represent a new understanding upon which new things can be created.

Let’s detour for a minute to think about Elon Musk.

Elon is an innovation tzar. I believe history will look back at his achievements and put him in the innovation hall of fame, alongside the greatest industrialists in human history.

He has revolutionized several industries to create multiple billion dollar companies in a row: PayPal, SolarCity, Tesla, SpaceX, and potentially HyperLoop, The Boring Company and/or Neuralink.

Some attribute his success to his pain tolerance. Watch this video when he tells the story of 2008 if you want to experience his pain first hand. But Elon himself talks about a different thing which might be the source of his genius:

“I do my favorite thing which is apply physics first principles. It’s like the best tool possible.”

Elon is a famously powerful learner, but what he does with that knowledge is what we can learn from him, in turn.

Innovator Logic = First Principles Reasoning

The way Elon thinks through problems was brilliantly described in The Cook And the Chef by Tim Urban, and from that article (which is compulsory reading for all humans) comes this perfectly simple and powerful diagram that captures the first essential skill for innovators.

 

In a simple graphic, Tim has nailed it. This is first principles reasoning in a nutshell, and what I’m calling Innovator Logic.

You see, human knowledge is vast, but it’s not all correct. Much of what we call normal life is built on faulty assumptions which are hidden in plain sight. And much of the future is waiting on the other side of those hidden and incorrect assumption being corrected.

If you want to watch Elon talking about first principles reasoning himself, there are good videos here and here.

The disciplines of research, history and science (among others) will add content to the body of human knowledge. Those disciplines will deepen our understanding and extend what we know.

First principles reasoning also extends human knowledge, but in a different way. It’s the process of correcting faulty assumptions embedded in established knowledge, which when corrected will open new pathways for innovation and growth. This is a learnable skill by any aspiring innovator, and is essential for creating genuinely new things.

How do you apply Innovator Logic?

There are tons of articles simply identifying first principles reasoning as a good thing to do, but few explain exactly how.

First principles reasoning requires adopting a scientific, physics-style of logic. The Wikipedia definition is this:

A first principle is a basic, foundational, self-evident proposition or assumption that cannot be deduced from any other proposition or assumption.

The bottom line is you have to keep peeling back assumptions until you get to the bedrock of observable, undeniable truth. It’s a questioning process: it’s about asking why 50 times until you’ve scraped back every possible judgement, assumption, belief, opinion or assertion — and found a bedrock of undeniable fact and truth. Then, it’s the process of rebuilding new logic pathways that are equally robust.

This is an intensely difficult exercise, plagued by the insidious challenges presented by cognitive biases, and the raw intellectual horsepower it can require to think through a complex puzzle in fresh way.

Most of us don’t do this. Instead, as Elon says, we reason by analogy. There is this common phrase in startup land that illustrates the point: The “X-for-Y”. Facebook for doctors. Uber for satellites. Amazon for marijuana. It’s the same as saying “just like X, but different”. The logic says “it worked for them, it’ll work for us, but differently”. X-for-Y is a useful communication tool, and might be a useful way to tell your story, but you’ve still got to show how you’ve moved the body of human knowledge forward.

Let me illustrate with a real-world example.

Here we go folks. Let me illustrate with a real world problem that hasn’t been solved yet. It’s a simple problem to describe, but deceptively complex to solve.

Problem: Waiting in line at a drive-thru.

Put your hand up if you hate that? Who doesn’t? Who on earth wants to be stuck in line behind 7 cars while trying to get a snack or coffee? Also, there is mysteriously at least one car every time which has 9 kids in it and an overwhelmed parent on the verge of murder trying to get them some ice-cream so they’ll shut up for a minute. Yes, that parent is facing a melted disaster in the back seat, so extra napkins please.

23 minutes later, you get to the pick-up window, then drive away and realize they have messed up your extremely simple order and you have to either (a) go around to endure the ordeal again, or (b) drink someone else’s blueberry mocha latte.

Why on earth has this not been solved? When you look at the category of quick service restaurants, no-one really has. Drive-thru have not changed much since they were invented in the 1930’s.

 

Some attempts have resulted in perfectly inadequately HALF-solutions. For example, McDonalds has introduced wishbone designs where two cars can order at once, but they are in fact more confusing than helpful. Starbucks has added TV screens where you can watch the person taking your order, but you’re probably just watching someone make a mistake in front of you. Some places have two lines, one for cash and one for mobile order-ahead. Nothing is the perfect solve, which is a signal that the hidden assumption is deeper in the logic tree than people in the industry are looking.

So what’s the transformational question? Where is the logic-flaw that is holding the whole category back from eliminating all the frustrations and pain points we always experience.

It’s this: drive-thrus have to deal with cars in series

There must be a line, because that’s the only way drive-thrus work. It’s how they’ve always worked.

The idea of serving cars in series, once upon a time, was a fundamental truth. In the first decades after drive-thrus were invented the only way to place an order was to a human, who communicated directly with other humans to fulfil that order. Today we have new technologies. So can we say the original assumptions that led to the original designs are still true?

If you dig into this problem, it doesn’t take long to find the logic-flaws hidden beneath the conventional queue+box+window design.

  • Do we need to order at a window? No, we can order on phones.
  • Do we even need to order on site? Nope, apps can let you order ahead.
  • Do we need to read a menu board? Nope, we have them in apps.
  • Do we need to hand cash to a person? Nope, we now have credit cards and electronic payments through our increasingly sophisticated smartphones.

The emergence of smartphones has made the old truths irrelevant.

From a first-principles perspective, the fundamental truth is this: the ‘maximum non-avoidable wait time’ is equal to how long it takes to fulfil an order after it is received. 

Any wait beyond that is either a function of (a) the customer’s behavior, (b) operational inefficiencies, or (c) physical drive-thru layout.

For coffee, that maximum non-avoidable wait time is how long it takes to pour the cup; for a burger, it’s the time it takes to grab it from the heat-stand and put it in a bag. We are talking a measure of seconds, not minutes. And definitely not 23 minutes.

The solution: a drive-thru design where vehicles are served in parallel. 

All drivers should be able to place their order, wait the maximum non-avoidable time, and pick up their order. This should be valid at all times, no matter how busy the store is.

To achieve that, many operational factors need to be changed. Sounds hard right? Not really. At first principles, all operational factors are a function lot-size, which in turn determines store layout which in turn has a huge impact on efficiency. Expertise in the industry says lot-size is hard to change, but it’s clearly not impossible to change, and is easily changed for new stores.

When you look at the McDonalds drive-thru wishbone design, it’s clear they got it half right. They have parallelized the point of order, but the rest of the experience is still in series. A fully parallel experience does not yet exist, but would fundamentally eradicate the wait time for drivers.

So now imagine a conversation between two individuals: one is arguing in favor of a in-series design, and the other for an in-parallel design. Who is going to win? The in-series design stands no chance.

The economics are important too. How much does it cost? How will customer behavior change? Each of these pieces are solvable. If you can solve a huge pain point for customers with a truly new experience … in this case a drive-thru with consistent wait times measured in seconds … it’s pretty hard to lose money. In fact, do it successfully and the competition becomes redundant.

The payoff: criteria for good ideas.

Innovator Logic unlocks the next step in the innovation process: coming up with the ideas themselves.

However the next step is NOT brainstorming: it’s the process of creating criteria for ideas that are built on the new logic tree. Good ideas will comply with that set of criteria, not just impress a room of people brainstorming together.

This is another thing Elon has done brilliantly. The best example is HyperLoop. Elon was “disappointed” that California was going to blow billions of dollars on ($68.4 billion in fact) on a project that was inferior to existing solutions, like flying. So before setting his mind and a small team of engineers on an alternative solution, he established some criteria for what an alternative mode of transport should be.

He decided that it should be:

  • Safer (than planes or cars)
  • Faster (than planes and trains)
  • Lower cost (both per ticket, and to build)
  • More convenient (than airports)
  • Immune to weather (unlike planes)
  • Sustainably self-powering (because Climate Change will kill us)
  • Resistant to Earthquakes (unlike trains)
  • Not disruptive to those along the route (through noise or land consumption)

A solution that achieved all these things was the only way to truly solve the problem. Any idea that didn’t comply with all of the criteria was not good enough. Then they created the Hyperloop.

 

The Hyperloop Whitepaper is essential reading to see the thinking in its full glory.

A big watchout: ‘expertise’ can get in the way.

It’s important to add that Innovator Logic is unlocked by a powerful belief: not everything we think is true, is actually true. That belief embodies a sense of intellectual skepticism that opens up the type of questions that let you dig deep into the logic tree that other people just consider to be right. (Update: Mark Suster just wrote a great post on this exact topic).

From Elon’s Reddit AMA:

One bit of advice: it is important to view knowledge as sort of a semantic tree — make sure you understand the fundamental principles, ie the trunk and big branches, before you get into the leaves/details or there is nothing for them to hang on to.

If you view knowledge as a ‘semantic tree’, and believe you most likely do not not have the full tree or correct tree, the continual journey is to build it out in the most correct way.

I have rarely seen this attribute in the corporate world.

Most of us are taught (in school etc) to build on the knowledge of others, not challenge or question it. With the added inertia of 20 years of experience and expertise, and it’s even harder for someone to get out of their own way and see things in a new light. The trick is to accumulate knowledge, not expertise. The latter has an element of self-assuredness, a little bit like “I know this topic, better than others”. It’s the type of self-view which closes our minds. Innovators need to do the opposite: question the knowledge of others, but most importantly their own. We have to constantly search for things that are demonstrably always true while keeping our eyes peeled for any information that shows us a better truth.

Some concrete action steps.

Innovator Logic requires a special mindset and the skillset for sharp logic.

Raw intelligence is really helpful by the way. Elon might be one of the smartest men on the planet, and it’s hard to argue he isn’t the smartest when it comes to applied intelligence.

For the rest of us mere mortals, there are things we can do.

1. Agree to the mission. If the team does not agree that first-principles reasoning is necessary for a breakthrough, it’s pretty hard to get there. Instead, the team will spin, focusing on what is subjectively convincing, instead of objectively true.

2. Visualize the logic. Logic is built in trees, and seeing it on a whiteboard or in a mind-map is really helpful. This is a MUCH more powerful exercise than brainstorming with sticky notes. Map out the logic instead.

3. Draw on diversity within the team. There is no telling where a fresh perspective is going to come from. No two people have lived the same year of life. Experience and expertise is irrelevant. Flaws that exist within conventionally accepted ‘truths’ can be observed and corrected by anyone. I have seen the most junior people on teams unlock huge problems simply with a powerful question driven by their own innate curiosity.

4. Avoid the temptation to jump to conclusions. That’s the whole problem in the first place. The world is inevitably built on assumptions. When using first-principles reasoning, you have no choice but to break those assumptions down. The ever present risk is missing assumptions, or adding new ones along the way.

5. Stay VERY open to new information, but be careful of experts. Often they are experts in the past, not the future. One way to think of expertise is that it’s knowledge inertia. Just like large companies trying to pivot, some experts will have an extremely hard time pivoting away from what they know. Inside some of that knowledge are the assumptions to fix.

6. Make sure you have enough time. Some innovation constructs are time-sensitive. This caps how deep into the logic tree you can go. The goal is to get to the ‘atoms’. This means peeling back the layers of established knowledge and rebuilding from the ground up. If you’re pressed for time, you might not get to the ground floor.

7. Validate with the scientific method. It has stood the test of time for a reason. When you rebuild the logic tree, you need to do it with evidence. This means your new logic steps are in fact hypothesis, and are not valid until they cannot be disproven. Only then can you say something is ‘always true’. Finding facts which cannot be disproved is the litmus test for valid new logic, and the final insulator against adding new cognitive flaws and biases into a logic tree that is not necessarily correct.

 

Truly new things are glorious.

I hope this was helpful. I know it’s a long post; it comes from my experience on the innovation battlefield. The future belongs to the bold, but change is made by those who think problems through in a new way.

Happy hunting!

Let me know what you think. Leave a comment or connect with me on Twitter.

 

 

 

Subscribe for Updates
Stay up to date with each new article as they're posted
Kosta Stavreas

Innovation Strategist & 2x Entrepreneur