Visible questionscomparable measured questions
Customer questions
Metric anatomy
Show the numerator and denominator—not just the percentage.
Different metrics answer different questions. They are intentionally not interchangeable.
Answers mentioning the brandcomparable successful answers
Successful answers
Answers citing own domaincomparable successful answers
Successful answers
Current metrics use comparable, successful answers for active questions in the confirmed project context. Each ratio shows what was counted and the total it was compared with. A percentage without that total is incomplete.
Different questions need different denominators.
Current metrics use comparable, successful answers for active questions in the confirmed project context. Each ratio shows what was counted and the total it was compared with. A percentage without that total is incomplete.
Question visibility
Active questions with at least one current answer mentioning your brand ÷ active questions with at least one current successful answer.
Two providers mentioning you for the same question still count as one visible question.
Provider coverage
Successfully completed question/provider pairs ÷ (all active questions × the effectively released providers in your plan).
A stopped or unstarted pair remains in the expected scope. A smaller accepted batch cannot make incomplete coverage look complete. Without an active matrix, the labelled fallback uses current successes and recorded exclusions.
Brand mention rate
Current comparable successful answers mentioning your brand ÷ all current comparable successful answers.
This counts answers, not distinct questions. It can differ from question visibility even for the same sample.
Answers with sources
Current comparable successful answers with at least one valid source URL ÷ all current comparable successful answers.
Several citations in one answer count once. A citation confirms a returned reference, not what the page says about your brand.
Own domain as source
Current comparable successful answers citing your canonical domain or a subdomain ÷ all current comparable successful answers.
An answer with external sources does not necessarily cite your website. Several own-domain citations in one answer count once.
Competitor coverage
Comparable questions with at least one answer mentioning a confirmed project competitor ÷ all comparable questions.
Observed alternatives, publishers, directories and sources remain separate. An unconfirmed name is not a confirmed competitor.
Answer mentions
The number of current comparable provider answers mentioning your brand or the named confirmed competitor.
One question can produce several answer mentions. The visibility comparison uses a separate count of distinct questions.
Comparability
Two points belong on one trend line only when the context still matches.
Question, company and website context, language, country, market, provider, API surface, release, model and methodology must match.
Compare⇢
Missing evidence
Unknown is not zero.
Failed, unavailable or non-comparable evidence remains excluded. An older success is not silently promoted when the latest eligible attempt failed.
Failed, unavailable or non-comparable evidence never becomes a zero visibility result or an answer without sources. It remains an explicit exclusion. Provider coverage also shows the planned work that has not completed.
Evidence chain
From raw answer to a later verification.
A useful action remains linked to the measured gap. Marking work complete records execution; only a later comparable measurement can verify a changed signal.
Important boundary
A website audit and an AI measurement are different checks.
The bounded website crawl checks public technical signals and makes no AI-provider request. Its score cannot prove an AI mention or recommendation.
Follow the evidence before drawing a conclusion.
Open the original question, provider answer, sources and measurement date behind a result. Technical details help explain the evidence; they do not change its meaning.
What belongs in the current sample?
Nalqerio selects the latest eligible terminal attempt for each active question/provider pair. A failed latest attempt stays an exclusion; an older success is not silently promoted. Historical, imported, noncanonical or disabled-provider records cannot silently count as current proof.
When are two measurements comparable?
The question, company and website context, language, country, market, provider, API surface, release, model and measurement methodology must match. Changed context can prevent a comparison. The first result is a baseline; it is not a trend or proof that an action worked.
What does measurement confidence mean?
These labels describe sample depth, not statistical certainty. Moderate requires at least 10 comparable answers across 5 questions. Strong requires at least 30 answers across 10 questions and at least 2 measured providers. All other samples are limited; an available but unmeasured provider adds no evidence.
Is website readiness the same as AI visibility?
No. A website audit checks public signals such as HTTPS, titles, metadata, company markup, crawler access, sitemap, canonical URL and viewport. The bounded website crawl makes no AI-provider request. Its technical score stays separate and does not guarantee a recommendation.
When is an action successful?
An action links a measured gap to a relevant page and a concrete improvement. Marking it complete records work done. Only a later comparable measurement can verify an improved signal; it does not prove the action alone caused the change.
How are API evidence and usage recorded?
A measurement records the question, language, country, provider, model, request context, time, outcome and returned sources. Provider-reported usage is retained when available. Cost estimates use versioned prices and remain estimates; the provider invoice is authoritative. Local analysis is identified when it makes no AI request.
Start 14-day free trial
What you can try yourself
Try it for free
Your own project, three customer questions and one AI snapshot in total across enabled providers.
Your 14 days start with the first project. Afterwards, saved results remain visible; new measurements and changes pause. No paid subscription is created.
You make the website changes.
Nalqerio provides evidence, prioritised actions and drafts. It does not publish content to your website. Re-tests are part of Starter and Pro; a better position is not guaranteed.
Keep monitoring afterwards
Recurring measurements, targeted re-tests and reports are included in Starter and Pro. Both plans include four complete measurement cycles per billing period.
Compare plans
Important boundary
A useful sample has clear limits.
OpenAI Responses, Gemini and Perplexity are separate API measurements. They are not screenshots or identical reproductions of consumer chat interfaces. Answers can vary with wording, model, time, market and available sources. Automatic extraction may be imperfect; keep the source available for review.
Nalqerio does not promise rankings, recommendations, traffic, market share or revenue.
Sources
See the evidence behind each number
This demo shows sources and original answers. Unknown data remains unknown; a mention is not a guaranteed ranking.
Explore the product tour →
Demo view · sample data, not a customer measurement
NALQERIO
Explore your next step
Understand the product, interpret the evidence and choose a suitable plan.