First principle approach in building startups and products

Published in

The Product Design Blog

10 min readFeb 5, 2020

Why did I write this article?

Recently I did a couple of polls to figure out general awareness and understanding of first principles. First poll was on twitter, most of these users are tech savvy, a lot of them work in tech companies building products and services. Some are students or working at established corporate. Second was done on a telegram community (Creators of Products) which mostly has folks from startup world — Product managers, founders, engineers and designers.

Most people weren’t aware of First principles, while those who were aware, 2/3 of them felt they had a first principles mindset
Around half of the folks in this poll were unaware of the concept.

This post attempts to introduce readers to the concept of first principles in the startup and product creation space.

Definitions

Firstly, what does Wikipedia say:

A first principle is a basic proposition or assumption that cannot be deduced from any other proposition or assumption.
In philosophy, first principles are from First Cause attitudes and taught by Aristotelians, and nuanced versions of first principles are referred to as postulates by Kantians.
In mathematics, first principles are referred to as axioms or postulates.
In physics and other sciences, theoretical work is said to be from first principles, or ab initio, if it starts directly at the level of established science and does not make assumptions such as empirical model and parameter fitting.

Secondly, Elon Musk is considered to be a firm believer of first principles and here is his definition and understanding

At this point, you should have a basic understanding of what does first principles mean in theoretical terms.

How are first principles relevant for building products?

The true aim of startups and product management is to solve problems for their users by building and iterating through products. This is a difficult problem because unlike anything we studied in school or college, this includes a lot of uncertainties. Founders and product creators have to live in a world where the number of choices are almost infinite and the resources are scarce. This forces them to decide at a high speed on a daily basis in the multi variable system they operate.

First principles thinking helps them steer through this jungle of decisions with some predictability.

The rules of first principles thinking in layman terms can be written as -

Reason everything. This is the breaking the problem statement part. Keep asking ‘Why’ till you reach the most fundamental truth about your business, your user and your product
Avoid Assumptions wherever and whenever. Don’t start problem solving with assumptions. Assume only when all possible sources of information available are finished
Create a culture within your team where progressive questioning and reasoning are appreciated

Let’s look at the three rules in some detail:

Reason Everything, Ask questions

Creating stuff is fun. When you build something and it works, it’s an unmatched feeling. But the sense of achievement and happiness comes with a bias. Creators, at times, feel that since they have built something, people will use it. This is not always true.

First principles approach tells us to ask these essential questions -

For existing products, Why would people move from what they are currently using to your product? What is the compelling reason? Is the product changing their lives significantly? Why would they pay for your product?
For new products, Why would people change their current behaviour to using your product? If they didn’t know you would they still use the product?

Just asking these questions is not enough, the next step to go deeper into these question and break down the problems statements behind these questions into smaller, more predictable, solvable chunks (This is where you can see the parallels with first principles in physics or maths, for those interested, here is a blast from past from calculus classes). Instead of hoping that your product will just become successful, you will be able to see smaller, more solvable problems and it will be a bit easier to attack them.

From Giphy

Here is a small list of some important questions that you will need to look at —

Market

Is the market large enough?
Have we reached out to the right kind of people?
Have we reached out to enough people?
Is the market ready for the kind of product we are launching?
Is the product for a large stagnant market or a small but growing market?
Is the market too mature for the product?

Product

Is this even solving a pain point for them?
Are users able to use for the first time without guidance?
Does it work all the time, in the best possible way?
Does it delight the users, for them to use it again and to bring other users?
Do users love it or like it?
How would life be if the product didn’t exist anymore?

Communication

Are we able to communicate what we want to solve?
Do they understand what it does?
Are users excited about sharing feedback?
Internally, is everyone in the company solving for the same thing?

Pricing

Does my expected segment want to pay for this?
Is there a different segment that would pay for this product?
Do users think its too costly or too cheap?
Are users using the product only for promotions?
Is price the only differentiator for our users?

Trust

Do users trust the product / startup?
Is our NPS getting better?
Are social media trends healthy?
Are more people happy with us or angry with us?
Do we give users what we say we will offer?

Execution

Can we test this hypothesis without spending a month of development?
Does building this feature really impact the key results we are aiming for, how?
How will the conversions from this fancy new feature be more than the current product flows?

These are only some examples on breaking down the problem statements into as many questions as possible. Some of these questions are before building the product, some during the product and some after launching. For different kinds of sectors and markets, these will change. What will not change is that building products purely on imagination will have a lesser chance of winning compared to building for solving fundamental real life problems. There is enough literature on internet about how to build a product and how to listen to your users, so I will not go deeper into that. For starters, one can go through all the videos in this lecture series by Y combinator.

In summary, I am a strong believer of form over function! And first principles provide validation for that. Do what is required, solve a problem for users and try to create predictability.

And the two questions you need to ask every day is -

Why would new users try our product?;
why would existing users keep coming back?

From Giphy

Avoid Assumptions

Most startups begin on assumptions. We assume the size of markets, how big the business can become, we will hire a large team and raise a fund. We also assume that users will love the product, they will bring more users, they will even pay for it now and in future. For a lot of startups, all the above happens. May be for a few it even happens exactly as per assumption, but mostly the assumptions only help with teams execute and move forward and are hardly on point with the reality of future.

What first principles teaches us is that even the most complicated problem can be brought down to solvable parts. The problem with that is that the process of breaking it down into such small part is an extremely complex, tiring, mentally exhausting and sometimes irritating.

Imagine a setup where the board or the CEO says that this quarter we will Grow 10x and will bring down costs by 5x. Is this a bad goal? Hell, no? This is a great goal for any company and if it can achieve this they will become the poster child of how to run businesses. But the complex part of this is — well, doing it!

Folks who have never had the taste of such responsibility can sometime risk taking up this goal with full enthusiasm, rigour and belief with the assumption that — “I am so smart I can do anything”, however, this can end up like the Dothraki sequence in the final fight sequence in that dark episode of Game of Thrones.

From Giphy

Unpredictability and unpreparedness for adverse results hurt more than bad ideas and ugly designs.

Another general approach towards humongous goals is jumping to ideas and assuming that ideas fix everything. Prima facie, this looks like the right thing to do and sometimes it works as well since it’s easy to build consensus on copying something that a competitor has done or something that sounds fancy. But this is akin to giving some generic medicine to patients without assessing the symptoms and figuring out the root of the problem.

First principles approach makes one create a possible world view of what would happen if certain things are done, or if certain other things are done. Loosely saying, this is somewhat similar to how doctor strange imagines the possible end game scenarios but in real life doing that in a practical, objective and measurable way.

From Giphy

Also, those who have tasted hyper growth at startups sometimes feel the bias that since their products in past have seen hockey stick growth, they will be able to solve every problem in the same way. But what generally take a company from zero to say, 100k users generally wont take it from 100k to 100mil users. And mostly, if it worked for one product doesn’t necessarily repeat for others.

So, should one never assume? Nope, this should not be misinterpreted as never assume. Building startups is a game of continuously testing various hypotheses and mostly without a lot of information on hand. The analogy here is from man vs wild. When the access to comfortable food options are limited you get frugal and you observe sharply. In the product building space in 2020, information is food. In absence of information, try to find out, within and outside your setup, if similar hypotheses have been tested in other situations. What happened when someone else tried to do what you are assuming will work. What worked for them, what didn’t work for them, was execution at fault or was market not ready. It is critical to not repeat mistakes that have already been done. This is the core of the concept of failing fast.

When given an incredible target like 10x growth and 5x cost reduction, build out possible scenarios where this can happen. Figure out opportunities that will help you achieve it. If you directly start by answering the questions — ‘Building what features will help you achieve this’, you are skipping the training sessions before the big match. Even Kohli and Messi don’t do that.

Culture of First principles thinking

First principles approach is not something you will learn in a workshop or a college. In my experiences, I have seen people get this by one of the two ways

By working in high accountability environment. This forces people to solve problems by first principles to bring maximum predictability and thus preparedness. High predictability doesn’t mean guaranteed results but it means less rework, better movement towards the aspired goals even though it is mostly in an uncool and unsexy manner
By working with leaders with strong first principles approach. People learn first principles by force in this case by the continuous questions of the leaders. And since such leaders practice this all the time, they keep getting better at it. This pushes the teams to think much deeper and harder on the problems they are working on and even better, it makes them more accountable and comfortable with saying things like — ‘ I don’t know’, ‘This is all the information we have’, ‘There is no evidence to support that the feature you are suggesting will work, please suggest the expected metrics movements on these suggestions’.

A true first principles approach team can only be formed if everyone is the team feel extremely comfortable in asking questions. This can happen when the leaders trust their team to think and own product and business problems as much as they do. There are detailed studies in this area, and this is known as a theory Y approach as well. In absence of this level of trust, the first principles approach can be a facade for randomly questioning everything that the team does. This is generally not very useful for the team or the product. In such a setup, it’s better to set expectations upfront that the company needs to be top-down, leadership tells and teams execute. It’s not right or wrong, different setups and different personalities operate better in different ways.

Another great practice to bring in first principle culture in the team is by using a common language for goal setting that everyone aligns with. In my experiences of using variety of these tools, finally I have found love with OKRs (read this book to learn most things about OKRs, a detailed OKR blog is in drafts as well). But depending on the size and stage of the product these tools can differ. What is important is that

a. The goal posts shouldn’t move too fast and too often and

b. There is transparency in goal setting process.

From Giphy

Goal setting is a primary, but underrated, problem in startups and should be treated as one. Imagining it as an overhead and only for corporates can lead to tickling down of similar culture in all other aspects of the startup. Culture is not what’s written on the walls but it is what people do everyday, if goal setting is done loosely and randomly, expect rest of the activities like consumer research, product development, delighting the consumer, figuring about sustainability will be done in the same way.

Lastly, make sure that first principles are not used as a shield for lack of movement and creating a culture of analysis paralysis. The beauty of product building in startups lies in the agility and that shouldn’t be slowed down at any cost. These are not contradictory concepts at all and should not be seen like that too.

From Giphy

The views shared in this article are personal.