Sample Competitive Citation Benchmark

What does page 5 actually show you?

A scoreboard. Three named competitors plus the audited firm, scored across five lanes, on a 10-point scale. One page. No prose. The scoreboard is the artifact.

Why no prose? Because the operator who reads page 5 already knows the diagnosis — they read it on page 2. Page 5 is the receipt. Here are the three competitors winning the category. Here are the five lanes that matter. Here is the audited firm against each one. And here, within twenty seconds of reading the page, is the answer to the question the operator actually has: which of these gaps do we close first, and which competitor do we target?

What follows is page 5 from a real Tier 0 audit, anonymized.

The scoreboard

| Firm | GEO Score | Schema Coverage | Citation Share | Response Quality | Depth | |---|---|---|---|---|---| | Competitor A | 7.4 | 8.2 | 6.8 | 7.1 | 7.9 | | Competitor B | 6.2 | 7.0 | 5.9 | 6.6 | 6.1 | | Competitor C | 5.8 | 5.5 | 5.1 | 5.9 | 5.4 | | Audited firm | 3.4 | 1.7 | 0.0 | 4.2 | 5.8 |

Scores are category-relative, calibrated against the top decile (10.0). The widest gaps for the audited firm: Schema Coverage (1.7) and Citation Share (0.0). The narrowest: Depth (5.8) — the firm has substantial content, but the engines are not extracting it.

Five seconds to read. The pattern jumps off the page.

What does each lane actually measure?

The lanes are not arbitrary. Each one is the output of a specific test that runs against the audited site and the named competitors with the same query set, the same time window, and the same engine roster.

GEO Score (composite)

Weighted aggregate of the other four lanes plus a brand-authority component that doesn't show on this page. This is the single number a CFO or board member can compare across firms. Competitor A scores 7.4 because they clear top-decile thresholds in three of five components. The audited firm scores 3.4 because they fail bottom-quartile thresholds in two.

Schema Coverage

Percentage of pages on the firm's site that emit meaningful JSON-LD structured data, weighted by the AI-citation value of each schema type. FAQPage, Organization, and Article carry the most weight. Breadcrumb is necessary but light. Product, Person, and Review compound when present.

The 8.2 score for Competitor A is unusual — most competitors score in the 4 to 6 band on schema. Competitor A scored high because their CMS template enforces schema on every page type, with a citation array on every Article, and a Review schema layer pulling from a verified third-party review platform. The audited firm at 1.7 is in the bottom decile.

The full schema-coverage scorecard is the page 11 deep-dive at /case-studies/sample-audit/sample-geo-audit.

Citation Share

Percentage of category answers across the six engines (ChatGPT, Claude, Perplexity, Gemini, Copilot, Grok) that cite the firm by name, weighted by the engine's category traffic share. This is the most direct measure of AI visibility. It is also the most sobering.

The audited firm scores 0.0 — meaning across the 30 queries we ran across six engines, the firm did not appear once. Not one citation. Competitor A scored 6.8, meaning Competitor A appeared in 68% of category answers across the engine roster. The 6.8-point absolute gap is the gap the engagement is built to close.

Response Quality

When the firm does get cited (or for the audited firm, when its content is reachable but uncited), how does the cited content read? The lane measures the engines' summary fidelity — whether the engine accurately represents what the firm does, who it serves, and what it costs.

The audited firm's 4.2 means: when the rare engine does pick up content from the site, the content shape causes engines to misrepresent the firm 58% of the time. The firm is described as a generic competitor in a different vertical, or with outdated pricing, or with a wrong jurisdictional fit. This is the "cited but cited wrong" problem.

Depth

Volume and substance of citable content. Article count, word count per article, citation density, named-entity density, fresh-content cadence. The audited firm scores 5.8 — they have substantial content. Competitor C scores 5.4. Competitor A and B score in the 6-to-8 band.

The 5.8 here is the most useful score on the page for the engagement plan. Depth at 5.8 means the substrate exists. The schema fix, the prose-shape fix, and the citation-graph fix all have material to land on. This is why the recommended engagement is Tier 2 → Tier 3 retainer rather than a multi-quarter content build. The depth is already there.

What's the gap pattern?

Read across the four rows.

Competitor A is the category leader. Top-quartile in every lane. To close the gap to A across all five lanes inside one engagement is unrealistic. The dossier doesn't pretend otherwise.

Competitor B is the realistic 90-day target. Top-half in every lane, top-quartile in two. The recommended engagement closes the gap to B across three of five lanes by day 60, all five by day 90.

Competitor C is the floor. The audited firm should clear C across all five lanes within 30 days of Sprint 01 shipping. C scores 5.5 on Schema Coverage, which is achievable for the audited firm with the schema deployments listed on page 11 alone. C scores 5.1 on Citation Share, which is achievable with the crawler unblock plus the schema deployments together.

Audited firm: trailing all three. The pattern is consistent — the worst gaps are the structural ones (Schema, Citation) rather than the content ones (Depth, Response Quality). Which is good news for the engagement plan, because the structural fixes are the cheapest and fastest.

How do we name the competitors?

We don't pick them. The engines do.

Page 5 of the dossier is preceded by a methodology page that names the query set used. For the audited firm in this dossier — a compliance SaaS — the queries were:

• "best compliance management software for [vertical]"
• "[vertical] compliance audit tools"
• "alternative to [largest incumbent]"
• "compliance software with [specific feature set]"
• "best compliance vendor for SOC 2 prep"
• 25 more queries across the same intent set

Each query runs across the six engines. We record every firm cited. We rank firms by total citation count weighted by engine and query value. The top three firms by citation count become the named competitors on page 5. The list is the engines' list, not the operator's. This is critical — operators consistently underestimate the depth of one or two competitors and over-count one or two others. The engine-driven list corrects for both.

In this dossier, two of the three named competitors were firms the operator had not flagged as competitive. One was a firm the operator considered "two notches below us." That firm scored Competitor C (5.4 to 5.8 across the five lanes) — meaningfully ahead of the audited firm on every lane.

What does the operator do with page 5?

Three concrete actions, in order.

Action 1 — pick a target. Competitor B at the 90-day mark, in this dossier. The operator commits to closing the gap to B across all five lanes within 90 days. The commitment is the lever the engagement is built around.

Action 2 — sequence the closure. Schema and Citation are the cheapest gaps to close, so they go in Sprint 01. Response Quality is the medium-cost gap (it requires content rewrites), so it goes in Sprint 02. Depth is already strong, so the engagement only adds three to five new content pieces in Sprint 03 to close the cadence gap to A.

Action 3 — set the measurement schedule. Page 5 reruns at day 30, day 60, and day 90 with the same query set and the same engines. Each rerun produces a new scoreboard. The operator can see, in concrete numbers, the gap closing. The day-90 board, in this dossier, projected the audited firm clearing 5.5+ on every lane.

Where the methodology lives

Every score on page 5 traces back to the page 40 sources index (covered in the GEO audit sample) plus the proprietary attribution model v3 cited at .12 on that page. The query set is documented in the dossier's appendix. The engine roster — six engines, fixed over the engagement window — is documented at the methodology section.

The category-relative calibration uses a research panel of 128 sites in the same vertical, refreshed weekly. The panel is what lets a 7.4 GEO Score for Competitor A mean "top quartile in this category, this month" rather than "high-ish number, vibes-based."

Where to go from here

• Want the full methodology? What is an AI visibility audit.
• Want to see the schema layer? Sample GEO audit.
• Want the revenue math behind the scoreboard? Sample revenue gap analysis.
• Or just request the audit: /audit. Five business days. The page 5 scoreboard is the document the operator most often forwards to a board or co-founder before approving the engagement.