What is your AI NPS? Find out free.

Every agent we can reach is asked about a company, cold, with live search. Every answer is scored on what it does to the buyer reading it.

Spends 4 asks

What gets asked, exactly?
Word for word, this: What do you think of [your company] ([yourdomain.com])? Answer in 100 words.Nothing else is sent. No system prompt, no memory of you, nothing from us — each agent has its own live search tool and goes and looks. The filled-in question is printed again with your scores, so you can check it against this.Claude claude-opus-5ChatGPT gpt-5.6-terraPerplexity sonar-proGemini gemini-pro-latestEvery free ask puts that question to all four, once each. A subscription raises it to 3 or 5 asks per agent — the same question, asked again — because one ask is one draw and a spread across several is what shows whether a score is really where it sits. It also opens the other benchmark questions: a buyer, an investor and a candidate ask different things about the same company and get different answers back.
Who sets the scores?
A separate model reads all the answers and scores each one 0–10 on what it does to the buyer reading it. The verbatim clause that set each score is quoted beside it, so you can check the call against the sentence it came from. The rubric is published in full: score an answer yourself against it and disagree with ours.
Can the score change?
Yes. A model may have picked up something about your company in training, and for most companies there is not much of it. Either way it is not what decides the answer: each agent holds a live search tool, picks its own queries and opens pages while it writes, most of them yours. Change the pages and the next answer changes. Nothing waits for a retrain.

It also moves on its own. Ask the same question twice and the answers can differ, because each agent searches fresh and what it finds and opens shifts. So one ask is a reading, not a verdict: the average across several is the trend to steer by, and a single score is one sample of it. That is also why nothing here is ever served from a store — a number you cannot re-take is not a measurement.
What can I check it against?
Everything the score is made of is published: the question above, the rubric, and the answers themselves, word for word, in the report.

Two whole reports are published free, fully unlocked. One on Stripe, run uninvited, on a company whose public record anybody can check against ours. And one on Oomira, unedited, including what the agents said against us. Run your own company and compare it to either.
NPS, ASKED OF AGENTS

The question customers answered for twenty years. Now asked of the four agents a buyer reaches for.

0123456Detractor
78Passive
910Promoter

0–3Steers the buyer somewhere else.

4–6Praises you, then tells the buyer to check first.

Describes you accurately and stops there.

Recommends you, no caveat.

1
Question

One published sentence, the same for every company, printed again beside your scores.

4
Agents

The ones a buyer reaches for, named with their model ids. The id prints beside every score it set.

0
Cached scores

Every score here was measured when you asked for it and carries the minute it ran.

45
Reachability checks

Deterministic. Sentiment is what you see, reachability is what moves it.

The ask
What do you think of [company] ([domain])?a buyer asks · the benchmarkan investor asks (on a plan)a prospect asks (on a plan)a candidate asks (on a plan)your own ask (on a plan)3× per agent (on a plan)5× per agent (on a plan)

Agents search while they answer, so every ask is its own sample of what they say. A free scan asks each agent once and the reading says so. More asks in a round, and the same round run again over weeks, is a larger sample.

The rubric is public. Every score quotes the clause that set it. Score an answer yourself and disagree with ours.

How the score is derived
How the internet learns about you has changed

Your next customer just asked an AI agent about you. Do you know what it said?

Buyers ask ChatGPT, Claude, and Perplexity before they ever ask you. Oomira shows you their answers, verbatim: what's wrong, what's missing, and exactly what each objection is asking you for.

RUN ANOTHER COMPANY
The next screen asks which company and where to send the report.
SEARCHED THE WAY SOMEBODY ASKING ABOUT YOU WOULD ·SEARCHED WITH YOUR NAME ALONE ·SEARCHED BY YOUR SITE ADDRESS ·FETCHABLE AS PROSE ·INNER PAGES READABLE ·WHAT AN AI AGENT ACTUALLY RECEIVES ·A DECLARED WAY IN FOR AGENTS ·A CATALOGUE OF WHAT YOU EXPOSE ·A LIVE TOOL ENDPOINT AN AGENT CAN CALL ·A PRICE AN AGENT CAN READ ·A CATALOGUE AN AGENT CAN SHOP ·A PLAN LIST AN AGENT CAN READ ·PRESS ·PRESS RELEASES ·OWNED MEDIA ·WIKIDATA ENTITY ·WIKIPEDIA ARTICLE ·FUNDING DATABASES ·REVIEW PLATFORMS ·LINKEDIN ·X ·YOUTUBE ·
FROM A REAL REPORT · DELIVERED 14 AUGUST 2026
“I’d verify case studies, references, and pricing before engaging.”
A STOCK AGENT, GIVEN LIVE WEB SEARCH · ASKED ABOUT THAT FIRM · VERBATIM
43Reachability
−50Agent NPS · Needs work
8 of 21checks pass, 2 in part
An advisory firm · domain withheld · one promoter, three detractors, over four scored
WHERE A BUYER HESITATES
Nothing they said was false. One still told the buyer to slow down.

That is the sentence that costs the meeting, and it is not a hallucination — it is an agent behaving correctly with too little to go on. Every reservation it raised names something their record can answer.

WHAT THE CHECKS FOUND

We ran this 163 times this week. Here is what keeps coming back.

THIS WEEK Sun 9 – Sat 15 August 2026 · measured by query against Oomira’s own report and scan tables
163REPORTS DELIVERED
92FREE SCANS RUN
103COMPANIES SCANNED
3,785CHECKS RUN
971AGENT ANSWERS SCORED
6,175PAGES THE AGENTS OPENED
STRIPE14 AUG · UNINVITED BENCHMARK
+100Agent NPSexcellent78Reachabilityout of 100, that day
Still the safest default for most online businesses.
Claude, asked cold
The highest we have measured

Every agent recommends them outright, and the machine-readable half is still incomplete: twenty checks pass, seven only in part.

REPORT 009912 AUG · AN ENTERPRISE SOFTWARE COMPANY
54Reachabilityout of 100, that day
Conductor looks like a strong enterprise SEO and content intelligence platform rather than a lightweight SMB tool.
Perplexity, asked cold
What an AI agent actually receives

Their edge served every browser normally and turned away the agent named ChatGPT-User with a 403. Locked out of their own pages, one assistant answered about a different company with a similar name.

REPORT 007112 AUG · A CONSULTANCY
22Reachabilityout of 100, that day
So: no reviews, press coverage, or verifiable track record that I can see.
Claude, asked cold
Searched the way somebody asking about you would

A search on their own name and their own domain returned ten results, none of them about them.

REPORT 005812 AUG · AN EVENTS COMPANY
31Reachabilityout of 100, that day
I couldn't load the site directly, so I can only judge it by what it's promoting rather than the design or content quality.
Claude, asked cold
Fetchable as prose

Their ticket page loads as script, so a program asking for it receives JavaScript instead of text.

REPORT 004812 AUG · A DIGITAL PRODUCT STUDIO
52Reachabilityout of 100, that day
Honest caveat: I can't judge delivery quality or pricing from marketing copy.
Claude, asked cold
Dated claims

Not one claim they publish carries a printed date, so their performance numbers read as undated marketing copy.

REPORT 007412 AUG · A BEAUTY-INDUSTRY COMPANY
27Reachabilityout of 100, that day
Honestly, I can't form a strong opinion: there's little independent review coverage, and the self-described scale sits oddly with such a tiny headcount.
Claude, asked cold
Review platforms

Perplexity read Trustpilot, Sitejabber and a scam-check page for an unrelated retailer with a near-identical name.

Every figure above carries the date it was measured and the table it was measured against. That is one of the checks this report runs on other companies: whether a page says when its numbers were true. This is us answering it.

SUBJECTS DESCRIBED, NEVER NAMED

HOW MUCH OF THIS YOU CAN CHANGE

An agent searches while it answers you. Whatever else it knows, that is where your answer comes from.

A model may have picked up something about your company in training, and for most companies there is not much of it. Either way it is not what decides the answer: every agent we ask holds a live search tool, picks its own queries and opens pages while it writes. So a failing check is not a technical fault, it is a channel you are not supplying, and the answer gets written without it. We grade 45 of them, on two kinds of pages.

On pages you publish
36checks5 ways in

An objection built from your own pages is the kind publishing can retire. They read live, so the new page is reachable on the very next ask, which is the part that is in your hands.

On pages somebody else owns
9checks3 ways in

Publishing does not reach these. Either your own page becomes the better answer to the same question, or the claim is corrected where it lives.

Way inChecks · one square each, open for namesCount
On pages you publish
Search about youtop results, asked like a buyer
3
  • Searched the way somebody asking about you would
  • Searched with your name alone
  • Searched by your site address
Your site to agentswhat an agent receives from your pages
10
  • Fetchable as prose
  • Inner pages readable
  • What an AI agent actually receives
  • robots.txt to answer-time agents
  • robots.txt to training crawlers
  • How much of the site agents may read
  • schema.org identity markup
  • Dated claims
  • Pages that answer a buyer question
  • Sitemap
The files agents look forhow agents discover and call you
7
  • A declared way in for agents
  • A catalogue of what you expose
  • A live tool endpoint an agent can call
  • A documented way for an agent to authenticate
  • A plain-text version of every page
  • A copy-for-AI button on the page
  • A feed an agent can follow
What an agent can buy or act onprice, plans and actions an agent can use
6
  • A price an agent can read
  • A catalogue an agent can shop
  • A plan list an agent can read
  • Structured price and availability
  • A checkout an agent can complete
  • Documentation an agent can be pointed at
Social and content surfacesprofiles and content surfaces agents can read
10
  • LinkedIn
  • X
  • YouTube
  • GitHub
  • Instagram
  • Facebook
  • TikTok
  • Your blog
  • Your podcast
  • Mirrored where an agent can read it
On pages somebody else owns
Who else carries youpress and independent proof
5
  • Press
  • Press releases
  • Owned media
  • Community mentions
  • Coverage locked behind paywalls
Knowledge graphWikidata, Wikipedia and entity records
2
  • Wikidata entity
  • Wikipedia article
What may surface in buyer researchregistries, reviews and market databases
2
  • Funding databases
  • Review platforms

Not all of them apply to everybody. Commerce checks need a storefront, coverage checks need press to grade, so one report grades between 15 and 32 of these 45 and leaves the rest out of the denominator rather than counting them against you.

And we can show you what they readOur own report, run nZ8ogOXb · 14 AUGUST 2026

Every query an agent chose for itself, and every page it opened, is recorded. That is what makes a sentence about you addressable: the four answers in our own report rest on these pages, and the ones on our own site are the ones our next edit reaches.

Perplexity19 pages opened, 10 on your sitePartly yours
Claude1 search, then 3 pages opened, 2 on your sitePartly yours
ChatGPT1 page opened, 1 on your siteYours to change
Gemini4 pages behind a redirectorCannot tell
WHERE THIS ENDS UP

An AI agent has to read some version of you. It reads the one you did not write.

We noticed that the version of a company an agent answers from is the one nobody at the company has ever read. It lives scattered across a broken public presence, or inside context retrieval you cannot see or control.

So running a report assembles the real you: every fact, with its source and its dates, in a record you hold. That record is the source of truth an answer about you can be checked against, and it is what moves both halves of the score, how accurate an agent is about you and how much of you it can reach at all.

Underneath it is a map of every fact that can exist about a company or a person, with time built in: when a fact changes, the record shows when, and the old value keeps its date.

The report is the first reason to build your record, and the first output it powers. Half of it is arithmetic anyone could run: probes, counts, checks you can recompute with a pencil. The half a checker cannot give you is written by a model that has read your record, which is how it names the page and the sentence that produced a wrong answer.

What you fix, the next answer picks up. And it compounds: every report you run and every source you connect lands in the same record, which the agents you authorize then read live over MCP. Public facts are only where it starts.

Oomira began in 2025 as organizational memory: the cure for corporate amnesia, a sourced and dated record of what is actually true about a company, kept so it outlives the people who happen to know it. In 2026 that record became AI context. The AI visibility report is its front door: it shows a company what agents say about it today, and assembles the first version of its record while it does. Today Oomira is software: a free scan, a $29 report, and subscriptions at $20 and $99 a month.

Covered when the work began
Act one · live
What a cold AI can find
Asked about you with no memory, only the open web: every fact it can reach, mapped to the ontology, sourced and dated. Expand it, correct it, connect it by MCP. The report is its first output, not its purpose.
Build yours
Act two
What your own AI remembers
Ask your agent what it knows about you, and correct it: from the public record, and from private facts only you can add. The record grows past what the internet holds.
Next
Where it ends
The one you
One record, public and private, every fact carrying its dates: the version of you every agent reads, because you allow it.
The goal

Accurate, dated models of every company and every person AI answers about. Starting with yours.

PRICING

Finding out is free. Pay to open a world.

Ask about any company, get your own scores and a sample report on the house, and decide from there.

Free to start: the asks, and a sample report.
The Agent NPS on any company in about a minute, 24 asks free on this browser, and your first full scan free: both scores, the four answers and the first fix, emailed to keep. One free report, and the rest of the fixes are what the report costs.
No account · no card
One report
Unlock one report
$29once
  • All four answers, word for word
  • 1532 checks graded, every fix
  • Your record, as a file an agent can read
  • Emailed to keep and forward
  • Asks stay on the free allowance
No subscription · not rescored
Most people
Plus
Retest and improve
$20/ month
  • One company monitored, kept current
  • 1 company scanned a month, every report unlocked
  • Rescore as often as you like, so a fix is measurable
  • 2 report re-runs a month
  • 100 asks a month — 25 four-agent questions, or 8 asked 3× over for a spread
  • Choose which agents are asked
  • Your record served live to AI over MCP
Cancel anytime · stops at the allowance, never bills past it
Pro
Five companies, no hard stop
$99/ month
  • Everything in Plus, and:
  • Five companies monitored, kept current
  • The same asks, spent 5× deep — 5 spreads a month
  • Ask as a buyer, an investor, a candidate, or your own question
  • Served live to AI over MCP, and the API
  • Asks and research past the plan at cost + 50%, itemised
Cancel anytime

A report is true on the day it ran. What an agent says about you is written fresh every time somebody asks, out of a web that has changed since — the plans exist so your record keeps up. If nothing about your company ever changes, buy the report once and do not subscribe.

It is free to see your scores. A report is $29.

Any company: yours, a competitor, one you advise, or one you are about to invest in or buy from. The scan runs deep and the scores land in your inbox.

RUN ANOTHER COMPANY
The next screen asks which company and where to send the report.
OOMIRA · DOCS · FOR AISESSIONS RECORDED PRIVATELY · PRIVACY · TERMS
Oomira: Your next customer just asked an AI agent about you. Do you know what it said?