Marginal Science: Product Analytics, Experimentation, and Pricing
We produce pricing, economic, and statistical analyses for your business.
We price products, design product experiments, and unlock what your data tells you about customer demand and behavior.
We teach you and your teams how to fish, too. We provide training on how to run experiments better and teach new causal inference and econometric tools.
We can:
- Be another set of eyes on an internal data, pricing, or economic analysis. A solid, low-cost way to get to know us.
- Improve your pricing using your data.
- Design an online experimentation program, and provide training for product managers and marketers starting out running their own experiments.
- Figure out where your conversion funnel performs well and where it doesn't.
- Dive deep to find out what's behind declining metrics.
- Learn what's driving your success. This is critical. It's tempting to not worry about success. But we should. We need to understand why business is good. So we can double down.
- Write the code for economic, statistical, or machine learning models. We can develop packages for your teams in essentially any language or framework. Reach out to learn more.
- Be plumbers. Write data pipelines and develop an easy to navigate database.
Our Principles
- Think from first principles. There are no best practices. Start from the beginning. Work out the solution step-by-step. No shortcuts.
- Write down what we want to know. Find out what we do know. Bridge the gap.
- To analyze means to cut. We write terse, direct reports relevant for making the important decision our clients face.
- No gloss. We include zero pictures of smiling people in suits in our reports. We produce facts and statistical/economic analysis, and document them in plain text. We design our work product for decision makers, not salesmen.
- Iterate. The best solutions are a process. An initial solution to try, a means to test whether it worked, and a plan for how to iterate towards the best solution.
- Care.
Contact Us
Blog Posts
To get a sense of how we think, check out some of our blog posts (warning: some posts include math):
- Losing Is The Value of Experimentation
- Build A Strong Experiment Culture Not A System of Experimentation Laws
- Use Decision Theory To Choose Significance Level For Experiments
- Does It Matter That Online Experiments Interact?
- In Defense of Statistical Significance.
About Us
Who is Marginal Science?
Marginal Science is a consulting company from Zach Flynn [LinkedIn]. He has a PhD in Economics and experience working on pricing, experimentation, and product analytics in the tech industries with companies like Udemy, Amazon, and Fivetran and in economic consulting (Compass Lexecon).
Where does the name Marginal Science come from?
In economics, we think about what happens on the margin. What if the price were a little more or a little less? What if a few more customers heard about your product? Answering these questions is key to understanding what you should do.
Not thinking on the margin is a mistake.
If we only think about our current customers, we don't grow. We're not stretching ourselves to the frontier. We're too far inside the margin.
If we think about customers who wouldn't even consider our current products, we'll have to make decisions without leveraging our key advantage: data and experience from the business we've built so far. We're looking too far ahead. The key is to think on the margin. How can we improve what we've done? How do we take the next step?
Where is Marginal Science?
We are based in Chicago, U.S.A. We work anywhere with internet.