Column · method

Counting demand before you build

Most people check demand after they have built something. There is a cheaper order: count how many different ways people already ask the question, then decide.

2026-10-19 · InfoGap column

What we actually count

For each of 130 areas of a life — from survival and health to money, family, law and meaning — we send two seed questions to Google's public suggestion endpoint and keep the long-tail phrasings it returns. A node's figure is the number of distinct phrasings recorded that day.

Today that is 2,798 phrasings across 130 nodes. Not a market size. A breadth of language.

Why breadth matters more than a single keyword

If a topic returns one phrasing, people mention it one way — usually because the word is a proper noun, or the topic is thin. If it returns thirty, the topic has many entry points: different products, different situations, different intents.

Thirty entry points is what lets you write a page that answers one of them precisely instead of all of them vaguely.

The trap: breadth is not traffic

We do not convert phrasings into search volume, because the conversion is not honest. A node can return thirty phrasings and have almost no buyers; another can return eight and hold a real market. Breadth tells you how a topic is spoken about, not how many people are buying.

A cheap sequence

1. Pick five nodes that describe work you could actually do.
2. Record their counts for two days. One day is a point, two days is a direction.
3. Open the node with the widest spread of phrasings, and read the examples. They tell you what people are trying to finish.
4. Write one page for the phrase that appears most often in a form you can answer.

That is the whole method. No tool needed beyond the page and a text file.

map · column · data · pricing · contact