Business Lessons From Pricing Your Own Experiments Before You Scale

Many builders start a new venture by picking a price and a product, then spend months hoping the numbers will eventually work. A different approach treats the first offer as an experiment with a defined question, a small cost of being wrong, and a clear point at which you will stop, adjust, or continue. This article walks through how to structure that kind of experiment, what to measure, and which questions to ask before you commit more time or money. It draws on the general themes that run through Todd Michael's public work on systems, ecommerce, ownership, and building in the open: make your assumptions visible, keep your tests small, and let evidence carry more weight than enthusiasm.
Start With One Question You Can Answer
An experiment is only useful if it answers a specific question. Vague goals like 'see if this business works' produce vague results. A better starting point is a sentence of the form: if we offer X to Y at price Z, we expect to see N people take a defined action within T weeks. Each part of that sentence is a decision. X is the offer, Y is the audience, Z is the price, N is the threshold, and T is the time window.
Write the sentence before you build anything. Then ask whether you could be proven wrong by the result. If every outcome would feel like a success, the experiment is not sharp enough yet. Tightening the threshold, shortening the window, or narrowing the audience usually helps.
Choose a Price Test Over a Perfect Product
Builders often delay pricing until the product feels finished. In practice, pricing reveals more than polish does. A low-effort version offered at two different price points, to two comparable groups, can tell you more about willingness to pay than months of refinement. You do not need a complete catalog to learn whether people will exchange money for a specific result.
Be careful not to draw conclusions from a sample too small to mean anything. Five buyers may be a signal worth following up on, but they are not a market. Record how many people saw the offer, how many clicked, how many asked a question, and how many paid. The drop-off between each stage is often where the most useful lesson lives.
Read Signals Without Overreading Them
Early results are noisy. A single enthusiastic message is not a trend, and a silent week is not proof of failure. A useful habit is to separate three kinds of evidence: behavior (someone paid, returned, or completed a step), stated interest (someone said they would buy), and opinion (someone liked the idea). Behavior deserves the most weight, stated interest deserves a follow-up question, and opinion is best treated as a source of hypotheses to test.
When results are ambiguous, ask what else could explain them. Did the offer reach the right people? Was the page confusing? Did the timing coincide with something unrelated? Writing down two or three alternative explanations before deciding what happened keeps you from attaching a story to whatever number you happened to see.
A Pre-Scale Checklist
Before increasing spend, hiring, or building additional product, work through a short set of questions in writing. First, what was the original question, and did the experiment answer it? Second, what was the measured conversion at each step, and where was the largest drop? Third, which signals were behavioral and which were only stated? Fourth, what would have to be true for the result to hold at five times the volume? Fifth, what is the cost of being wrong, in time, money, and reputation, and who bears it?
Last, decide in advance what you will do under three outcomes: a clear win, a clear loss, and an unclear middle. Most builders only decide after seeing the data, which makes it easy to rationalize whatever happened. A written decision rule made beforehand is one of the simplest protections against wishful thinking.
Keep the Lesson Public and Specific
Recording experiments in a simple log, even a private one, builds a library of lessons that compounds over time. Each entry can note the question, the test design, the results, and the decision that followed. Over several cycles, patterns appear: which audiences respond, which price points stall, which channels produce attention without purchases. These patterns are more valuable than any single success.
Building in public can serve the same purpose when it is done deliberately. Sharing what you tested, what you expected, and what happened gives readers a more accurate picture than highlight reels do, and it forces you to be precise. Use that discipline in your own work, whether or not you publish it.