InfoGap / answers / industry demand
How do you tell if an industry has real demand?
One ratio does most of the work: the share of an industry's searches that are decisions rather than questions.
In one line (cite this)
According to InfoGap's industry-demand dataset (source: our own harvest of search-suggestion phrases, ten decision-type seed phrases per industry; definition: deduplicated search phrases per candidate industry, split into decision-type and browsing-type phrases; collected 2026-10-04, dataset generated 2026-10-05), 569 of 570 candidate industries returned phrases — 73,174 deduplicated phrases in total, of which 35,116 (48.0%) are decision-type.
Count decision phrases, not searches
The fastest filter is the decision share: the fraction of an industry's distinct phrases that look like someone about to act, not someone reading.
- 73,174 distinct phrases were collected across the 570 candidate industries.
- 35,116 of them (48.0%) are decision-type. Browsing or knowledge-type phrases account for only 1,816, or 2.5%.
- The ratio is not uniform, and that is the point: restaurants sit at 0.644, hotels at 0.598, hospitals at 0.551 — against 0.333 for big data.
| Industry | Decision share |
|---|
| Restaurants | 0.644 |
| Hotels | 0.598 |
| Hospitals | 0.551 |
| Big data | 0.333 |
The empty rows are rarer than you expect
Of 570 candidate industries, exactly one returned no phrases at all; the other 569 carried some signal.
- 569 of 570 candidates returned at least one search phrase.
- The single empty row was explosives, pyrotechnics and fireworks manufacturing.
- Size still varies widely: hospitals lead with 187 distinct phrases, 6.2 times the bottom entry, courts and procuratorates at 30.
Negative phrasing is countable, and it clusters
Negative queries are a small share overall, but they concentrate in a short list of industries rather than spreading evenly.
- 5,791 of the 73,174 phrases (7.9%) are negative.
- 50 industries carry 20 or more negative phrases.
- Hospitals and e-commerce lead that list with 30 each; restaurants follow at 26.
What people get wrong
The usual mistake is treating search volume as demand. This dataset never measures volume at all.
- Volume is not intent. Every figure here is a count of distinct phrases, not of how many people searched. A phrase list can be long and still be pure curiosity.
- A high phrase count is not a high intent score. Big data has phrases, but only 0.333 of them read as a decision, against 0.644 for restaurants.
- One empty row is not a verdict. Only one of 570 candidates came back empty. That is a result of this harvest, not proof that nobody wants the product.
- Negative phrases are not noise. 7.9% of all phrases are negative and 50 industries carry 20 or more — worth reading, not filtering out.
Frequently asked questions
What is the single fastest signal that an industry has real demand?
The decision share: the fraction of distinct phrases that look like a decision rather than browsing. Across all 73,174 phrases it is 48.0%.
How many industries were checked?
570 candidates, ten decision-type seed phrases each. 569 returned at least one phrase, for 73,174 deduplicated phrases in total.
Which industries have the strongest decision share?
Restaurants 0.644, hotels 0.598, hospitals 0.551, against 0.333 for big data.
Do negative searches matter?
5,791 of 73,174 phrases (7.9%) are negative, and 50 industries carry 20 or more of them.
Turn this into a reading of your own
The same method works on any industry you pick. Start from what people already say they want, then compare it with what you can actually deliver.
Get your chain →
Free, no account and no email needed to start. Your answers stay in your browser until you ask for a chain that follows you.