Applying First Principles Thinking to your Startup Idea to Enhance Impact

One thing that surprises me over and over is how many startup founders don’t use models and equations in their business. I always get strange looks when invariably I ask my favorite business questions: “So, what is the equation here? What variable are you hoping to affect by doing this?”Startup founders and startup people in general are all about “metrics” and being “data driven”. These things are not what I am talking about when I talk about “equations”. Metrics are inputs and variables on “equations” that model other operational and financial metrics.If you have a website that generates traffic that you hope to convert to customers then how many visits you generate in a month and how many visitors you convert to customers are both metrics. Some metrics examples include: 1000 Visits 20 New customers 2% Customer conversion rate 100 Canceled accounts 3% Churn rateNow, let’s build a simple equation as an example:

New Visits * ( Conversion Rate ) = New Customers

This is simple, and at first you could say that it is too simple to be useful. Now suppose that your goal is to increase the number of new customers you have over a given period of time. You can use this equation in two ways: 1) To filter ideas or actions 2) To generate new ideas or actions.

Choosing between actions:

Suppose you have a list of things you want to do: 1) Rebrand the company: I want to change my logo and name because I don’t think they’re good enough.
2) Go to a Startup Event: I can network with other entrepreneurs, learn from experienced people during talks and even meet some investors. 3) Start doing a paid ad campaign to drive traffic to my website. 4) Add a new cool feature to the product.

Now let’s get back to the equation:

New Leads * ( Conversion Rate ) = New Customers

Looking at your 4 possible actions which ones would move you in the right direction? The thinking here would go like this:1) Rebrand the Company: Is this going to increase my visits? I don’t think so. Will it increase my conversion rate? Maybe it could, but I don’t think so. Can this generate more customers outside of the equation? I don’t think it will right now.2) Startup Event: How is this is going to increase my conversion rate? I doubt that it would. How is this is going to increase my visits? Maybe I can meet an influencer that would promote my website to his or her audience, but this does not seem to be very likely to happen. Can this generate more customers outside of the equation? Maybe, if I sell to startups some of the people at the event could become my customers, but it is not likely.3) Paid Ad Campaign: Is this going to increase my traffic? Probably, yes. Is this going to increase my conversion rate? I don’t know, but depending on the campaign it could. Can this generate more customers outside of the equation? No.4) Add new feature: Is this going to increase my traffic? I don’t think so. Is this going to increase my conversion rate? I don’t think this feature is enough to make someone that was not interested in the product become interested. Can this generate more customers outside of the equation? No.This type of Socratic questioning will help clarify your options and make it easier to decide what is the right course of action. This type of questioning is not possible if you don’t have simple equations that describe your key metrics.

Generating new Actions

My favorite use of this kind of thinking is to develop an action plan from the equation. Let’s remember our example:

Leads * ( Conversion Rate ) = New Customers

You have a few options considering this equation: a) Maintain the Conversion Rate and increase Leads b) Maintain visits and increase the Conversion Rate c) Increase bothI have two levers to generate more customers: 1) Increase my visits: How can I drive more traffic to my website? What are the biggest sources I have today? Can I improve on them? 2) Increase my conversion rate: How does my conversion rate compare to the industry? What can I test to increase my conversion rate?Your questions should not stop at this level. For each key variable on your equation you should generate an equation that has this metric as output.1) Leads: what are the key components that drive my number of leads? One good example could be:

New Visits * Conversion Rate = Leads

Now you can do the same process thinking about ways to drive up your new visits and conversion rate.

The wrong reasons to not use equations

One criticism of this structured approach I hear a lot are:“You can’t model everything using formulas.” Yes, you can’t model everything but this is not a reason not to model what you can. The fact that you can’t use a fork to write does not mean that the fork is useless. There is no such thing as a silver bullet in business; context is king.“Formulas are too simple to capture the reality behind real business problems.” Equations simplify complex processes into core principles. Capturing the complexities of a business is hard work, but I argue that having a good enough model is better than deciding by gut feeling.

Examples of Equations

I love SaaS business because of the wealth of literature around models. It’s really easy to find widely known and used equations.One of the most vital formulas for you company is the LTV one:

(ARPA * Gross Margin%)/(Customer Churn Rate) = LTV

The LTV model captures so many interesting relations and properties of your business at once that it deserves an entire post just to unpack this single line. The LTV is a measure of how much one customer is worth on average and increasing your LTV is a good way to have a healthier business. I was talking to a fellow entrepreneur recently and he told me this:“I need to get better retention, I am totally focused on this right now.”After a few minutes of conversation I discovered that what he really wanted was to increase his LTV. The automatic advice on the industry is to increase your retention so that your Lifetime Value will increase. This is true, but it is not the only way of doing it. Don’t get me wrong, I am all for improving customer retention but it’s not a silver bullet to solving your LTV problem. Using models is a great way to test advice like “Improve your retention to get a better LTV.”Just looking at the LTV formula you can tell that there are three ways to improve your LTV:Increase your ARPAIncrease your Gross MarginsDecrease your Customer Churn RateIf your goal is to improve your LTV you now have three levers you can use. Maybe the solution will be decreasing your churn rate, maybe not. The important thing to have in mind is that by analyzing your equations you will understand all your options and make better decisions.

The case for using more equations

There are three main reasons to use more equations in your business: 1) Equations as planning tools Choosing what metrics you want to move and establishing goals is a good first step toward improving your business. Understanding what equations govern the metrics you want to move will give you more clarity on what levers you can pull to achieve your goals. 2) Equations as decision making tools Your resources are limited and saying yes to one thing means saying no to many. Prioritization is the heart of great execution. Using equations as a filter for your actions will help guide your efforts and help you avoid wasting time and resources. 3) Equations as communication and alignment tools Communicating and aligning your goals with your team is a constant challenge for any startup founder. If your team understands the basic equations involved in a particular decision or plan, the alignment and communication will be orders of magnitude better.

The case against using more equations

The biggest reason to avoid equations is that overusing this tool can lead to a too narrow view of things. This view can limit you and your company to a local maxima and in order to achieve a big breakthrough you need to think outside your current equations and create new ones.

Conclusion

Math is an elegant language to describe relationships between things. More than only metrics, thinking about models and equations or how the metrics influence each other is a powerful tool for running your business. Don’t let the simplicity of this tool fool you; the most profound laws of nature are simple as well.