The company, stated plainly
What Vector Index is, and what it is not
Vector Index is a company. It measures how AI assistants answer the questions buyers ask before they call a business, and it improves the pages and profiles those assistants read. It is not a vector database, an embeddings library, or the vector index data structure used in retrieval systems. Those share our name and they are not us.
Five of five engines read our own name as a data structure.
We sell measurement, so we measure ourselves and publish what comes back. Asked about “Vector Index AI Search Intelligence” by name on 26 September 2026, three engines explained vector embeddings, retrieval databases and semantic search, and two asked which of several technical meanings we meant. Not one of the five described a company.
That is the problem we are paid to find, and we have it. A name shared with a standard technical term is not a branded query a business wins by default, and no amount of good work reaches a buyer whose assistant cannot tell which kind of thing you are. This page exists to say which kind of thing we are, in words a person and a machine can both read.
Source. 5 questions on ChatGPT, Gemini, Claude and Perplexity through their APIs with web search enabled, plus Google AI Overviews read from Google results pages · one repeat per question · 24 of 25 answers returned, 1 search showed no AI Overview · sampled 26 September 2026 · stored verbatim. One repeat cannot produce a rate, and this does not.
What we do
One thing, in a fixed shape. We ask the questions a buyer would ask, on named engines, on a recorded date, and we keep every answer exactly as it came back, including the ones that failed.
- We sample. 25 questions, four repeats each, on ChatGPT, Gemini, Claude and Perplexity through their APIs, plus Google AI Overviews read from Google results pages for the market in question. Every run records the engine, the model, the settings, the date and the run number.
- We report what the answers did, not what we hope they mean. Whether the business was named. Whether one of its own pages was cited. Which sources the engine leaned on instead. Attempted and successful samples are counted separately, and every report states what was not measured.
- We write three change specifications. What to change, on which page or profile, and why that change follows from the evidence. In words, not code.
- We measure again at 90 days on the same panel, and say plainly that a change in answers is not proof of a change in customers.
How we measure, in full · What the $750 audit includes · Ask for a free sample
What we will not sell you
Each of these is sold in this category. We decided against all of them before we had a client, and the reason is the same every time: we cannot show you the evidence for it.
- A guarantee of rankings, citations or traffic. Nobody controls what a model writes, and anyone who tells you otherwise is describing a hope.
- An llms.txt file. It is a real proposal that no engine has committed to reading. We wrote up what we actually found.
- A Wikipedia page for a business that does not meet Wikipedia's notability standard.
- A visibility score out of 100. A single number hides the population it was counted over, which is where every honest finding lives.
- A blended hit rate across engines, questions and markets. Averaging five engines into one figure describes none of them.
The company
- Legal name. Vector Index LLC, an Ohio limited liability company. “Vector Index” in conversation, “Vector Index AI Search Intelligence” in full.
- Who runs it. Justin Ashcraft, founder.
- Where we work. Cincinnati and Indianapolis are the starting markets. The work is delivered remotely across the United States.
- Who we work with. Independent businesses selling considered purchases: home services, elective self-pay medical practices, funeral homes and flight schools.
- Mail. 7965 N High St, Ste 350 #208, Columbus, OH 43235.
- Email. [email protected]. There is no phone number yet, and we would rather say so than publish one nobody answers.
Implementation work is planned as a separate, priced scope and is not for sale yet. The one thing you can buy today is the $750 audit.
Questions people ask before they reply to us.
Is Vector Index a vector database?
No. A vector index is a data structure used in embedding and retrieval systems, and Vector Index is a company that measures how AI assistants answer buying questions. We do not sell a database, an embeddings library or search infrastructure of any kind.
Is Vector Index legitimate?
Vector Index LLC is an Ohio limited liability company with a published mailing address, a named founder and a single priced service. Every figure on this site traces to a stored sample with a date on it, and where we have measured nothing we say so rather than estimating.
Who runs Vector Index?
Justin Ashcraft, the founder. He sends every client deliverable himself.
Do AI assistants recommend Vector Index?
Not yet, and we publish the measurement rather than the ambition. On 26 September 2026 we put four non-branded questions from our own category to five engines, and we were named in none of the 19 answers they returned. We measure this on a schedule now, and the same dated method is the one a client's audit uses.
What does Vector Index charge?
$750, prepaid, for one brand, one main site and one market. There is also a free AI Answer Snapshot, which is a much smaller sample sent by email.
Does Vector Index do SEO?
No. We measure a different surface and we do not compete with a business's existing search work. The difference is set out in AEO vs SEO.