monitoraeo
AI Visibility Rankings · Methodology

How the rankings are produced

Every industry page at /ai-visibility is generated from a fixed methodology: 8 category-level queries against Google AI Overviews per brand, scored monthly, with a composite ranking that weights being-recommended over being-cited. Here are the specifics so you can audit the data yourself.

In one sentence

For each industry, we run 8 buyer-intent queries against Google AI Overviews, count how often each brand is named in the prose and cited in the source list, compute a 70% visibility + 30% citation composite score, and re-rank monthly. Same questions every month so the trend is comparable.

The 8 queries we ask, per industry

Identical question shape across every industry — only the category name changes. These are buyer-intent prompts that mirror how AI engines are actually used in research:

1. Best {industry}
2. Top {industry} 2026
3. Best {industry} for small business
4. Top {industry} alternatives
5. How do I choose the right {industry}?
6. {industry} pricing comparison
7. {industry} reviews
8. What is the best {industry} on the market?

Eight is the right number: enough to see real patterns (not just single-answer noise from one query), but small enough that the cost stays bounded at ~$0.03/refresh/industry. The shape mirrors the free-preview query set in our paid audits, so the methodology is consistent across products.

The scoring formula

For each brand in an industry, we compute two raw metrics:

visibility_pct = (answers_naming_brand / 8) × 100
citation_pct = (answers_citing_brand_domain / 8) × 100

Then the composite that drives the ranking:

visibility_score = visibility_pct × 0.7 + citation_pct × 0.3

The 70/30 weighting reflects an editorial choice — for most brand owners, being named in the AI's prose recommendation matters more than being one of many cited sources. If you weight them differently (e.g. 50/50 because you're optimising for citations as a backlink-equivalent), the ranking shifts but rarely dramatically. The raw visibility_pct and citation_pct are visible on every brand row so you can re-rank with your own weighting.

Refresh cadence

Every 30 days by default per industry. The cron worker picks up to 3 industries per 5-minute tick whose next_scheduled_refresh is due, runs the 8 queries, re-scores every brand, snapshots the previous state into history (drives the Movers & Shakers section), generates a fresh AI-written narrative + FAQ from the new data, and advances the schedule.

What's stored and what's regenerated

Per refresh, the system updates: visibility + citation percentages per brand, the composite score, rank ordering, per-engine breakdown (currently just Google AI), top cited sources for the whole category, and a fresh AI-generated narrative (3 paragraphs of observations + per-top-5-brand context + 4–6 category-specific FAQs). The previous state is snapshotted into monitor_industry_brand_history so the next refresh can compute Movers & Shakers.

How the brand list per industry is curated

We hand-pick ~20 brands per industry from public market-leader sources: G2 Grids for SaaS categories, Crunchbase top-funded lists, trade-press "best X" coverage. The intent is broad coverage of recognised leaders, not exhaustive coverage of every product in the space. If your brand is missing from a category you think you should be in, send us a support ticket with the industry slug and we'll review.

What we DON'T do

How to improve your ranking

The same way any AEO program works — better content, structured data, entity signals, earned mentions on the sites AI engines trust in your category. The 20-item AEO checklist is the practical playbook. The fastest single move is usually shipping FAQPage schema across your content library.

Audit your brand → Browse all industries

Related

Frequently asked

How often are the rankings refreshed?

Every 30 days by default — configurable per industry. The 'Refreshed [date]' pill on each industry page shows when the data was last regenerated.

What is the composite visibility score?

A weighted average of visibility (70%, % of answers naming the brand) and citation rate (30%, % citing the brand's domain). Reflects that being recommended in the prose is more valuable than being cited as a source.

Which AI engines are queried?

Google AI Overviews currently. The 5-engine paid-audit model (ChatGPT, Claude, Perplexity, Gemini) is planned for the public rankings later in 2026.

How are the 20 brands per industry chosen?

Hand-curated from public market-leader sources (G2, Crunchbase, trade press). Broad coverage of recognised leaders, not exhaustive. Missing brands can request inclusion via support.

Can rankings be gamed?

Not by us — we don't take payment for inclusion or position. Brands can improve their ranking the way any AEO program does: better content, schema, entity signals, earned mentions on sources AI engines trust.