The LBOX AI Industry Index
Find out how your brand and products rank in AI Answers
Who is leading AI answers in your industry, how you compare, and who you are really competing against.
Ranked across 256 industries, including the ones no other tool measures. Every ranking works the same way: the questions your buyers actually ask, put to the engines they actually use.
Tracked across the engines your buyers ask, through
Your brand is named in fewer than 1 in 5 answers.
Company rankings
Who gets named when a buyer asks an engine about suppliers, systems or compliance in your category. The position bars show where in the answer each company tends to land, first mention on the left. Being named fourth is not the same as being named first.
41 further companies in this industry appear in fewer than 5% of answers.
Sample data. Figures illustrative, company identifiers redacted.
Product rankings
Buyers rarely ask which company is best. They ask which system, which grade, which specification. These are the products named in answers to those questions, and the company behind each one.
A product can outrank its own manufacturer. That gap is usually the fastest thing to fix.
Sample data. Figures illustrative, company identifiers redacted.
The answer depends on who you ask
The same question put to different engines returns a different order. A company can lead on one and be absent from another, which is usually where the opportunity sits.
ChatGPT
Cites brand and trade press
Perplexity
Cites standards bodies
Google AI Overviews
Cites directories
Gemini
Cites manufacturer sites
Sample data. Figures illustrative, company identifiers redacted.
Who the engines pair you with
When a company is named, these are the companies named alongside it. This is often where a brand discovers it is being compared to someone it does not consider a competitor.
Sample data. Figures illustrative, company identifiers redacted.
What the engines are reading
The domains cited when answering questions in this industry. Owned sites appearing high is the clearest evidence that publishing works.
Sample data. Figures illustrative, company identifiers redacted.
Category context
Some categories produce named companies readily, others get answered generically with nobody named at all. Where your category sits tells you how much room there is to be the one that gets named.
Sample data. Figures illustrative, company identifiers redacted.
Where the gap is widest
Industries where buyers are asking and engines are answering without naming anyone. A wide bar to the right is unclaimed ground.
Sample data. Figures illustrative, company identifiers redacted.
How this is measured
Stated plainly, because a ranking nobody can interrogate is not worth publishing.
Buyer-intent prompts
Questions derived from real search demand in each industry, not invented. The prompts for every industry are published.
Every engine, same schedule
One prompt set is put to each engine on the same schedule, and results are recorded separately rather than averaged into a single number.
Repeated runs
Engines do not answer identically twice. Each prompt runs several times and the result is the average, so a ranking does not move without a reason.
Shown empty when thin
An industry without enough observations shows no ranking rather than an unreliable one. Coverage and last refresh are on every page.
Not seeing your company?
If your industry is covered and your company is not appearing, that is a finding rather than an error. Tell us the market you sell into and we will show you what the engines are saying.
Helpful to include:
- Your brand and the category you sell into
- The products or ranges that matter most
- Competitors you expect to be compared against
Sample data throughout. This is a design build, not a measurement. Industry names are real because they are the categories LBOX covers. Company identifiers are redacted and every figure on this page is invented.