The essays are general by design. The questions people ask me after reading them are not. Is this a network effect, or just scale? Which side should we subsidize? Why did replies fall when signups went up? The answers depend on the business, so I'm building an agent that starts from yours.

What it will do

  • Ask how your business works: who the sides are, how they find each other, what a good outcome looks like for each, and how you charge.
  • Tell you whether you have a network effect and where it lives: which side, which segments, and how strong it is likely to be. Sometimes the honest answer is that you don't have one, which is worth knowing too.
  • Walk through the parts that matter for you, such as density and tipping points, two-sided matching, ranking, trust, pricing and acquisition, with a link to the essay behind each point.
  • Finish with the few numbers worth tracking and the first analysis to run.

An illustration

An example of the kind of conversation it will have. Not a live chat.

You

We run a marketplace for home tutors in three cities. Signups grow every month, but parents keep telling us they can't find anyone good.

Agent

Before the totals, look at where parents actually search. A parent wanting a class 10 maths tutor only gains from tutors nearby who teach that subject at that level. How many tutors does a search like that return in a typical locality, and how many of them reply within a day?

You

In the older localities, plenty. In the newer ones, two or three, and most don't reply.

Agent

Then much of your growth is landing in segments that can't hold yet. Two things to check. First, reply rate by locality and subject: where supply is thin, parents leave before the next tutors arrive (Density beats size). Second, whether your ranking keeps sending parents to the same few popular tutors, who stop replying because they're full (When both sides have to say yes). Would you like to start with where to spend on acquisition, or with matching in the localities you already have?

Where it stands

It's in the works, and it's being built around the same principles as the essays. It is meant to work from a description of your business and the few numbers you're comfortable sharing, not your customer data. If you'd like to try an early version, send me a message on LinkedIn.