NIXUMA
NIXUMA

See how AI assistants recommend your business.

NIXUMA checks whether your business appears, which other businesses are named, and whether descriptions match your confirmed facts. You receive a written report and a correction plan.

Our method

We measure the answers, not your website. First, you confirm the facts of your business: hours, prices, services, location. Those confirmed facts are the ground truth for everything that follows.

Then we ask the questions guests ask, 30 per language, on the AI systems guests actually use. We ask each question three times on each system, in each language you choose, from English, Japanese, Italian, Chinese and Spanish. That is 450 answers per language, and up to 1,350 across three languages.

Answers per language
450
Answers across three languages
Up to 1,350

Why ask the same question three times?

An AI answer is not stored anywhere. It is generated fresh every time someone asks, and the same question can draw a different answer an hour later. One answer is an anecdote, not a measurement. Asked three times on five systems, one question returns fifteen answers per language, and the patterns show: the fact that is wrong every time, the name that appears only sometimes. Where repeated results move, we report a range, never a single confident number.

Why measure in more than one language?

The internet underneath the answers changes with the language of the question. In our second published study, the sources behind Japanese and English answers overlapped by only about a quarter, and for one system the two source sets had nothing in common at all. A business can be described accurately in one language and wrongly in another, and guests ask in the language they think in. Measuring one language shows you one version of your business. Your guests meet the others.

Why five AI systems?

ChatGPT, Gemini, Perplexity, Claude and Google AI do not read the same sources, and they do not treat them the same way. In the same study, one system carried fifteen sources beside every answer and used almost none of them in its text, while another leaned on official websites. Your guests are spread across all five systems. A check on one system measures one assistant's view. Five systems measure what your guests will actually hear.

How it works

  1. Measure

    Full Audit

    Check how often AI recommends your business, who else it names, and whether it gets your facts right.

  2. Fix

    Fix Pack

    Correct inaccurate details and create content that helps AI understand and recommend your business.

  3. Monitor

    Monitoring Scan

    Repeat the same tests to see whether recommendations increased and corrections held.

  4. Build visibility

    Use each scan to see what worked and what to improve next, so AI can recommend you more consistently.

Fix what the scan finds, then monitor again

Fix Pack

$189 One-time

Your audit found the problems. The Fix Pack produces the corrections, and applies them.

Six ready-to-apply deliverables, plus the Crawler Access Check, built from your audit's correction list and ordered by the damage each error does:

  • Corrected fact sheet.

    One reference document with your confirmed hours, prices, location, services and policies, written to be copied anywhere your business is described.

  • Structured data for your website.

    Ready-to-add markup that states your confirmed details in a format machines can read without guessing.

  • AI files for your website.

    An llms.txt written from your confirmed facts, ready to publish at your site's root, plus a check of your robots.txt with the exact lines to add if AI assistants are currently blocked from reading your site. It makes your confirmed facts available in the format proposed for AI systems, and it costs nothing to publish.

  • Crawler Access Check included.

    We test whether AI crawlers can actually reach your site, fix what blocks them, and test again.

  • FAQ copy.

    Clear answers to the customer questions your audit found answered incorrectly or inconsistently, ready to publish on your website.

  • Listing corrections.

    Ready-to-use wording for every listing that shows outdated or conflicting facts, each correction matched to the listing it fixes.

  • Content matched to real questions.

    Pages on your website written or reshaped around the questions your customers actually ask, built from your audit's own question set, so the page an AI reads answers the question it is asked. How much is written is agreed with you in the correction list.

Application is included for your website and Google Business Profile. With your written approval and revocable editor or manager access, we apply the agreed corrections and send you a change log. For third-party listings and sites you cannot grant us access to, you apply the changes with our written guidance and email support. We never ask for your password.

Buy Fix Pack

Monitoring Scan

$79 Per scan

Same questions, five AI systems

We repeat your Full Audit measurement using the same questions, languages, and number of runs.

  • See what changed since your audit.

    Compare how often AI recommends your business and which competitors it names.

  • Check whether errors came back.

    See which corrected facts are still reported accurately and which details AI is getting wrong now.

  • Compare ranges across repeated answers.

    See how much results varied and which changes deserve a closer look.

  • Get a one-page report by email.

    Delivered within seven days of the measurement.

A Full Audit is required first. We recommend scanning monthly. Your first scan is $59 if booked within 30 days of receiving your Full Audit report. No subscription.

Covers the same languages as your Full Audit. Extra languages are added at checkout (+$30 each) and chosen after purchase, from English, Japanese, Italian, Chinese and Spanish.