Citation Breadth vs Citation Depth: Which AI Authority Metric Should CMOs Care About?

Answer Capsule

CMOs should not choose between citation breadth and citation depth. They answer different questions, and using either one alone can produce the wrong publisher, PR, or distribution decision.

In the 283-market benchmark, 36,273 domains entered at least one eligible platform Top 100, but only 2,743 of them, 7.56%, reached a Top 10 in any market. Breadth identifies how widely a domain appears. Depth identifies how strongly it performs where it appears.

The distinction becomes obvious when domains with similar breadth are compared.

• TechRadar and CB Insights both reached the Top 100 in 109 markets. TechRadar reached a Top 10 in 65 markets. CB Insights reached a Top 10 in zero.
• NerdWallet and Alibaba both reached the Top 100 in 132 markets. NerdWallet reached a Top 10 in 97 markets. Alibaba did so in 11.
• Walmart, Consumer Reports, and Medium each reached the Top 100 in 145 markets. Their Top-10 breadth was 66, 25, and 5 markets, respectively.
• PCMag, OreateAI, and BusinessWire each reached the Top 100 in 91 markets. Their Top-10 breadth was 45, 5, and 0 markets.

Breadth therefore cannot be treated as a proxy for prominence.

The relationship is measurable across the whole corpus. Top-100 market breadth and Top-10 market breadth had a descriptive Pearson correlation of 0.717, but the rank-order relationship was much weaker, with a Spearman correlation of 0.306. Broad domains often accumulate more deep positions in absolute terms, but their relative ordering changes substantially when the question becomes, “How often is this source actually near the top?”

The same problem appears inside markets. Among the 260 markets with all six platform Top-100 portfolios available, the median market contained 310 distinct Top-100 domains across the six AI systems, but only 28 distinct domains appeared in a Top 10 on at least one platform. The median Top-10 union was only 8.75% of the Top-100 union.

For a CMO, the practical rule is:

Use breadth to understand coverage. Use depth to allocate attention and budget. Use both when deciding whether a source is strategically important.

A broad source may be useful because it touches many markets or platforms. A deep source may be more valuable because it repeatedly sits near the top in the one category that matters to the business. Neither metric proves editorial quality, audience value, recommendation influence, traffic, conversions, or future AI rankings.

Key Findings

Breadth and Depth Measure Different Forms of AI Citation Authority

Questions This Section Answers

1. What is the difference between AI citation breadth and AI citation depth?
2. Which AI citation metric should a CMO use to measure coverage versus prominence?
3. Why can a widely cited domain still be a weak source target in a specific commercial market?

Citation breadth answers where a source shows up.

Citation depth answers how prominent the source is where it shows up.

Those questions should not be collapsed.

This research uses several related measures:

These metrics describe different dimensions of the source ecosystem.

A domain can be broad but shallow. It may appear in dozens or hundreds of markets while rarely entering a Top 10.

A domain can be narrow but deep. It may appear in only one commercial category while ranking near the top across several AI platforms in that category.

A domain can also be broad and deep. Reddit, YouTube, Forbes, NerdWallet, and several large comparison or technology properties demonstrate that pattern to varying degrees.

The correct metric therefore depends on the decision.

If the question is, “Which domains have the widest footprint across our portfolio of markets?” breadth matters.

If the question is, “Which sources deserve investigation for a sponsorship, PR target, distribution partnership, or content placement in this category?” depth becomes much more important.

The parent study, AI Search Has a Tiny Universal Core and a Massive Specialist Long Tail, shows why global breadth alone misses much of the market-specific source ecosystem.

Only 7.56% of Top-100 Domains Ever Reached a Top 10

Questions This Section Answers

1. What percentage of domains that enter an AI platform Top 100 ever reach a Top 10?
2. How selective is Top-10 AI citation authority compared with Top-100 portfolio inclusion?
3. Does appearing in an AI platform Top 100 usually mean a source is a leading authority in that market?

Across the strict eligible market-platform portfolios, 36,273 distinct domains appeared in at least one Top 100.

Only 2,743 appeared in a Top 10 anywhere in the 283-market cohort.

That is 7.56%.

The other 33,530 domains, 92.44%, entered a Top-100 portfolio somewhere but never reached a Top 10 in any eligible market-platform portfolio.

This does not mean those sources are unimportant. Rank 20, rank 40, or rank 80 can still represent real citation visibility. A source can also serve a specialized use case that is not captured by a broad rank threshold.

But it does mean that Top-100 presence and leading-source authority are not interchangeable concepts.

The gap becomes even more relevant for marketing decisions because a global rank can accumulate many low-intensity appearances. A domain that appears at rank 50 to 90 across a large set of categories can look globally important while being less relevant to a category-specific publisher strategy than a specialist that ranks in the Top 5 on four or six platforms inside one market.

This is why the corpus keeps Top-100 breadth, Top-10 breadth, platform breadth, and response presence separate rather than blending them into a single composite score.

Breadth Is Correlated With Depth, but the Ranking Changes Substantially

Questions This Section Answers

1. Does a domain with broad AI citation coverage usually have deep citation authority too?
2. How strongly are Top-100 market breadth and Top-10 market breadth related across cited domains?
3. Can a global AI citation leaderboard reliably rank the best sources for a specific market?

Broad domains do tend to accumulate more deep positions in absolute terms.

Across the 36,273 Top-100 domains, the descriptive Pearson correlation between Top-100 market breadth and Top-10 market breadth was 0.717.

That sounds high, but it does not mean the measures are substitutes.

The descriptive Spearman rank correlation was only 0.306.

The difference matters.

Pearson correlation is influenced by the large absolute counts of a relatively small number of extremely broad domains. Reddit, YouTube, Google, Forbes, and other universal or near-universal properties create a strong positive relationship between “more markets” and “more Top-10 markets.”

Spearman correlation asks a different descriptive question: do domains maintain a similar relative ordering when ranked by breadth versus depth?

The answer is much weaker.

A domain can rise or fall dramatically once the ranking changes from “how many Top-100 markets?” to “how many Top-10 markets?”

This is exactly the failure mode a CMO should worry about. A leaderboard built only from breadth can identify important global sources, but it can also place a broad low-depth source ahead of a narrower source that is much more prominent in the advertiser's actual category.

These correlations are descriptive calculations from the frozen portfolio data. They are not inferential tests of a causal relationship.

Same Breadth, Very Different Depth

Questions This Section Answers

1. Can two domains have the same AI citation market breadth but radically different Top-10 authority?
2. Which websites illustrate why global AI citation breadth should not be used as a stand-alone publisher ranking?
3. How much can Top-10 authority vary among domains with identical Top-100 market coverage?

The cleanest way to see the distinction is to compare domains with the same Top-100 market breadth.

The TechRadar-CB Insights comparison is particularly useful.

Both domains entered an eligible platform Top 100 in 109 markets.

TechRadar reached a Top 10 in 65 of those markets. CB Insights reached a Top 10 in none.

That is not a small adjustment to the same story. It is a completely different authority profile.

The same pattern appears at 132 markets. NerdWallet reached a Top 10 in 97 markets, with a median Top-100 rank of 5. Alibaba reached a Top 10 in 11 markets, with a median rank of 47.

At 145 markets, Walmart, Consumer Reports, and Medium share identical Top-100 market breadth but very different depth profiles.

A leaderboard based only on Top-100 breadth would treat each breadth tie as roughly equivalent. The actual portfolio positions show why that would be misleading.

Some Very Broad Domains Were Almost Never Leading Sources

Questions This Section Answers

1. How many broadly cited domains never reached a Top 10 in any commercial market?
2. Which widely distributed websites had broad Top-100 citation presence but little or no Top-10 authority?
3. Can a domain appear across all six AI platforms and still have shallow citation depth?

Broad-but-shallow authority is not rare.

Among the 329 domains with Top-100 market breadth of at least 25 markets, 60 domains, 18.24%, never reached a Top 10 anywhere.

Among the 116 domains with Top-100 breadth of at least 50 markets, 8, or 6.90%, never reached a Top 10.

And 36 of those 116, 31.03%, reached a Top 10 in five or fewer markets.

Several examples stand out:

BusinessWire is a particularly useful example of why platform breadth also needs context.

It appeared in Top-100 portfolios on all six AI platforms and across 91 markets, yet it did not reach a Top 10 in any market in the strict portfolio dataset.

That does not mean BusinessWire has no value. It means “appears across all six AI platforms” is a coverage statement, not a depth statement.

LinkedIn makes the same point at much larger breadth. It reached the Top 100 in 248 markets but a Top 10 in only 28. Medium reached 145 Top-100 markets but only five Top-10 markets.

If a CMO is choosing a source to investigate for one category, these differences matter more than the raw market count alone.

Broad Domains Also Differ Greatly in Their Depth Conversion Rate

Questions This Section Answers

1. What share of a broad domain's Top-100 markets typically convert into Top-10 markets?
2. How often do domains with citation breadth across 50 or more markets achieve deep Top-10 authority?
3. Can a CMO use a Top-10-to-Top-100 breadth ratio to distinguish broad-and-deep sources from broad-and-shallow sources?

One way to compare breadth with depth is to calculate a simple descriptive ratio:

Top-10 market breadth / Top-100 market breadth

This is not a universal authority score. It is a useful diagnostic for asking how often a domain's broader portfolio presence translates into a leading position.

Among the 116 domains with Top-100 breadth of at least 50 markets:

• the median ratio was 14.55%;
• the 25th percentile was 5.76%;
• the 75th percentile was 28.34%;
• the 90th percentile was 48.05%.

The distribution is highly uneven.

46 of 116 broad domains, 39.66%, had a depth ratio below 10%.

77 of 116, 66.38%, had a ratio below 25%.

105 of 116, 90.52%, had a ratio below 50%.

A small group of domains is different.

These domains combine meaningful breadth with meaningful depth.

That is a different authority profile from a source that appears widely but usually near the bottom of Top-100 portfolios.

The ratio should still be interpreted carefully. A domain with only one market and one Top-10 market has a 100% ratio, which says nothing about global scale. Breadth and depth therefore need to remain visible side by side.

Narrow Sources Can Be Extremely Deep

Questions This Section Answers

1. Can a domain with AI citation authority in only one market still rank in the Top 10 across all six major AI platforms?
2. Which specialist websites demonstrate strong local AI citation depth despite minimal global breadth?
3. When should a CMO prioritize a one-market specialist over a globally broad publisher?

The specialist tail contains the opposite authority profile: low breadth and high depth.

Among domains whose Top-100 breadth was exactly one market, 61 reached a Top 10 on at least four AI platforms in that market.

26 reached a Top 10 on at least five platforms.

12 reached a Top 10 on all six platforms.

Examples include:

HearingTracker is the cleanest illustration.

Its global Top-100 market breadth is one. A global leaderboard would barely register it.

Inside Hearing Aids, however, its six platform ranks were 1, 4, 1, 2, 1, and 1 across ChatGPT, Gemini, Google AI Mode, Google AI Overviews, Microsoft Copilot, and Perplexity.

That is an authority profile that breadth alone cannot express.

This does not establish that HearingTracker is the best media buy, the most independent publisher, or the most influential source for recommendations. It establishes that in the observed Hearing Aids source portfolios, the domain had unusually deep cross-platform citation authority.

The companion article Why Niche Publishers Matter More in AI Search Than Global Citation Rankings Suggest provides additional specialist examples and publisher comparisons.

Inside a Typical Market, the Top-10 Source Set Is a Small Fraction of the Broader Top-100 Universe

Questions This Section Answers

1. How many distinct Top-100 and Top-10 citation sources does a typical commercial market contain across six AI platforms?
2. What share of a market's broader AI citation-source universe reaches a Top 10 on at least one platform?
3. Why should a CMO separate broad source discovery from high-priority source targeting?

The breadth-depth distinction also appears within markets.

Among the 260 markets with all six eligible platform Top-100 portfolios, the median market contained:

• 310 distinct domains across the union of the six Top-100 portfolios;
• 28 distinct domains across the union of the six Top-10 portfolios;
• a Top-10 union equal to 8.75% of the broader Top-100 source universe.

The middle half of complete markets contained 298 to 331.25 distinct Top-100 domains but only 25 to 31 distinct Top-10 domains.

That creates a natural two-stage process.

The Top 100 is useful for discovery. It identifies the wider source universe surrounding the market.

The Top 10 is useful for prioritization. It identifies the much smaller group that is leading one or more platform-specific portfolios.

The ratio still varies by market.

This is another reason not to apply one fixed “authority cutoff” everywhere.

Some markets have a relatively broad set of domains that reach the Top 10 somewhere. Others have a much narrower leading-source layer despite similarly large Top-100 universes.

The correct shortlist should therefore be market-specific.

Platform Breadth Is Not the Same as Market Depth

Questions This Section Answers

1. Does appearing on all six AI platforms mean a domain is deeply authoritative in the markets where it appears?
2. Why should a CMO report AI platform breadth separately from Top-10 market depth?
3. Can a source have six-platform coverage while rarely becoming a leading citation source?

Platform breadth creates another measurement trap.

A domain can appear in eligible Top-100 portfolios on all six systems somewhere in the cohort without being a leading source in many markets.

BusinessWire again makes the point. It had:

• Top-100 market breadth: 91 markets
• Top-100 platform breadth: 6 platforms
• Top-10 market breadth: 0 markets
• Top-10 platform breadth: 0 platforms

Medium reached all six platforms somewhere in the strict Top-100 portfolios, with Top-100 breadth of 145 markets, but reached a Top 10 in only five markets and on only three platforms.

LinkedIn reached all six platforms in the Top 100, but its Top-10 breadth was 28 markets and five platforms.

Contrast those with G2:

• Top-100 market breadth: 86
• Top-10 market breadth: 49
• Top-100 platform breadth: 6
• Top-10 platform breadth: 6

G2 is narrower by market count than LinkedIn, Medium, or BusinessWire, but much deeper across the markets where it appears.

For a CMO, “six-platform source” should therefore never be used as shorthand for “strong on six platforms in our category.”

The market and the rank still matter.

Breadth and Depth Should Be Used as a Matrix, Not a Single Score

Questions This Section Answers

1. How should a CMO combine AI citation breadth and depth when evaluating publishers or third-party sources?
2. What is the difference between broad-and-deep, broad-and-shallow, narrow-and-deep, and narrow-and-shallow AI citation authority?
3. Should AI citation breadth and depth be collapsed into a single weighted authority score?

A practical reporting model is a two-axis matrix.

Broad and deep

These sources appear across many markets and frequently reach leading positions.

Examples from the observed cohort include Reddit, YouTube, Forbes, NerdWallet, TechRadar, G2, Money.com, U.S. News, PCMag, and several category-specific comparison publishers.

They are useful when a company operates across multiple categories or wants source relationships that can matter in more than one market.

Broad and shallow

These sources appear widely but rarely reach leading positions.

Examples include CB Insights, BusinessWire, Medium, PR Newswire, Quora, and several general-purpose or distribution-oriented domains.

They may still be useful for discovery, credibility, referral, distribution, or other business goals. The citation data simply says their breadth should not be mistaken for deep Top-10 authority.

Narrow and deep

These sources have limited market breadth but strong local prominence.

HearingTracker, WATI, Wareable, BuilderOnline, ProvokeMedia, Lifeway Mobility, HouseFresh, Action Behavior Centers, and other specialists illustrate this profile.

For a focused category strategy, this quadrant can contain some of the most important sources to investigate.

Narrow and shallow

This is the largest part of the long tail. A domain may appear in one or a few Top-100 portfolios without strong rank, repetition, or cross-platform presence.

These sources should not be ignored automatically, but they require more evidence before receiving strategic attention.

The matrix is preferable to an opaque composite score because the tradeoff remains visible.

A CMO can decide how much the business values reach versus category concentration without pretending there is one scientifically correct weighting.

Which Metric Should a CMO Care About?

Questions This Section Answers

1. Should a CMO prioritize AI citation breadth or citation depth when allocating marketing budget?
2. Which AI citation metric is most useful for corporate reporting, publisher selection, and market-specific GEO strategy?
3. How should marketing teams choose different AI citation metrics for different business decisions?

The answer depends on the decision.

For budget allocation in a specific market, depth should usually carry more immediate weight because it reflects what the AI systems are actually using prominently in that category.

For enterprise coverage and portfolio planning, breadth becomes more important because a source that is deep in one market may not help across 20 business lines.

For publisher evaluation, both are necessary. Breadth describes the publisher's reach. Depth describes the strength of its observed authority where the advertiser cares.

For AI recommendation measurement, neither is sufficient. Citation authority is a source metric, not a recommendation metric.

That distinction is developed in AI Citation Authority vs AI Recommendation Authority: Why Being a Source Is Not the Same as Being Recommended.

Do Not Replace Breadth and Depth With an Arbitrary Composite Score

Questions This Section Answers

1. Should AI citation breadth and depth be combined into one authority score for executive reporting?
2. What information is lost when market breadth, platform breadth, and source depth are compressed into a single number?
3. How can CMOs compare AI citation authority without inventing arbitrary metric weights?

A composite score can be convenient, but convenience is not the same as validity.

Any combined score requires a weighting decision.

Should one Top-10 market equal five Top-100 markets? Should six-platform breadth outweigh a Top-3 rank on one platform? Should finance markets count more than consumer markets? Should a persistent source receive twice the value of a volatile source?

The study does not contain a scientific answer to those management questions.

The breadth-depth data demonstrates why the weighting matters. TechRadar and CB Insights have identical Top-100 breadth but radically different depth. HearingTracker has almost no global breadth but extraordinary local depth. LinkedIn has enormous breadth but a much smaller Top-10 footprint.

A single score would hide those profiles unless its components remained visible.

The safer executive design is to preserve a small set of independent KPIs:

1. market breadth;
2. platform breadth;
3. local source depth;
4. persistence over time;
5. portability across platforms;
6. recommendation authority measured separately.

Management can then decide which dimension matters for the current business objective.

The companion article Which AI Citation Metrics Belong on a CMO Dashboard? develops the reporting structure further.

Methodology

The analysis uses the frozen July through September 2026 cohort of 283 qualifying tracked markets, 876,605 retained AI responses, and 8,718,629 citation occurrences resolved to registrable domains.

The six consumer AI surfaces are:

• ChatGPT
• Gemini
• Perplexity
• Microsoft Copilot
• Google AI Overviews
• Google AI Mode

The strict platform analysis contains 1,655 eligible market-platform Top-100 portfolios.

A platform cell was eligible only when it contained at least 100 retained responses, at least 100 resolved registrable domains, and an actual ranked list of 100 domains. Thin or missing cells were excluded rather than assigned zero.

Within each eligible market-platform cell, domains were ranked by the number of distinct retained responses in which the registrable domain appeared. Repeated citations to the same domain inside one response counted once for ranking purposes.

For this article:

• Top-100 market breadth is the number of qualifying markets in which a domain appeared in at least one eligible platform Top-100 portfolio.
• Top-10 market breadth is the number of qualifying markets in which a domain appeared in at least one eligible platform Top-10 portfolio.
• Top-100 platform breadth is the number of the six AI platforms on which the domain appeared in an eligible Top-100 portfolio somewhere in the cohort.
• Top-10 platform breadth is the corresponding number for Top-10 portfolios.
• Depth ratio is Top-10 market breadth divided by Top-100 market breadth. It is used only as a descriptive diagnostic, not a composite authority score.
• Median rank is calculated conditional on the market-platform cells in which the domain actually appeared in the Top 100. It does not assign a synthetic rank to absent cells.

The Pearson and Spearman correlations reported above are descriptive calculations across the 36,273 distinct Top-100 domains. They summarize association between observed breadth and depth measures. They are not causal estimates and do not treat domains as independently randomized observations.

The six-platform market-union analysis is restricted to the 260 markets with all six eligible platform Top-100 portfolios. For each market, the Top-100 union contains every distinct domain appearing in any of the six platform Top-100 lists. The Top-10 union contains every distinct domain appearing in any of the six platform Top-10 lists.

Limitations

Several limits should remain visible when using breadth and depth commercially.

Breadth is a footprint, not a quality score

A domain can be broad because it is a publisher, marketplace, company site, social platform, search property, community, government source, press-release service, or another source type. Breadth does not establish editorial quality, independence, trustworthiness, or audience value.

Depth is not recommendation authority

A Top-3 citation source can be used as evidence while the AI system recommends a different company. Citation depth should not be interpreted as company preference.

Rank is portfolio-specific

A domain's rank depends on the market, platform, response denominator, observation window, and portfolio definition. A global median rank does not replace local inspection.

The study does not certify identical prompts across every platform and month

The historical dataset does not establish a fixed panel of identical prompts, prompt versions, model versions, countries, retrieval settings, and collection configurations across all observations.

The results describe observed source portfolios rather than controlled platform effects.

Platform breadth can come from different markets

A domain appearing on all six platforms somewhere in the cohort does not mean it appears on all six inside the same market. Local platform breadth has to be checked separately.

Top-10 thresholds are useful but not magical

Rank 10 and rank 11 are not separated by a natural law. Top-10 and Top-100 cutoffs are descriptive tools for showing differences in portfolio depth. Response presence, cited URLs, persistence, and local context should also be examined.

Business value is not measured here

The study does not measure publisher prices, referral traffic, conversion rates, brand lift, recommendation changes, sponsorship performance, or revenue. Those belong in commercial diligence or controlled campaign measurement.

Conclusion

Citation breadth and citation depth answer different questions.

Breadth tells a CMO how widely a source appears.

Depth tells a CMO how strongly that source appears where it matters.

The 283-market data shows why neither can substitute for the other.

Only 7.56% of domains that entered an eligible Top 100 ever reached a Top 10. Sixty domains with Top-100 breadth of at least 25 markets never reached a Top 10 at all. TechRadar and CB Insights reached the Top 100 in the same 109 markets by breadth count, yet their Top-10 breadth was 65 versus zero. NerdWallet and Alibaba both had Top-100 breadth of 132 markets, but Top-10 breadth of 97 versus 11.

At the same time, a one-market specialist such as HearingTracker could rank inside the Top 10 across all six AI platforms in its category.

Inside a typical complete market, the broader six-platform Top-100 universe contained about 310 distinct domains, while only 28 reached a Top 10 somewhere. The leading-source layer was a small fraction of the broader citation universe.

For CMOs, that leads to a simple operating model:

Breadth is a coverage KPI. Depth is a prominence KPI. Persistence is a durability KPI. Portability is a transfer KPI. Recommendation authority is a separate outcome.

The strongest AI search measurement program keeps those dimensions visible instead of compressing them into one number.

Related LLM Authority Index Research

• AI Search Has a Tiny Universal Core and a Massive Specialist Long Tail: How Citation Authority Fragments Across 283 Commercial Markets is the parent breadth and fragmentation study.
• Why Niche Publishers Matter More in AI Search Than Global Citation Rankings Suggest examines narrow but deep source authority.
• How to Find the Market-Specific Publishers AI Systems Actually Cite focuses on source discovery inside a defined market.
• How Publishers Can Build Category Authority Without Becoming a Generalist Media Site develops the publisher-side implications of narrow, deep authority.
• The Persistence-Portability Gap at Scale measures temporal durability versus cross-platform transfer.
• Top 25 vs Top 100 AI Citation Sources: Where Durable Authority Actually Lives examines portfolio depth as a stability dimension.
• The 2026 AI Commercial Citation Authority Index reports multiple authority dimensions without collapsing them into one arbitrary score.
• Which AI Citation Metrics Belong on a CMO Dashboard? translates the measurement framework into executive reporting.
• AI Citation Authority vs AI Recommendation Authority separates source authority from company recommendation behavior.

About LLM Authority Index

LLM Authority Index measures how AI systems cite sources, surface publishers, and structure authority across commercial markets. Its research separates market breadth, platform breadth, source depth, persistence, portability, and recommendation authority so that widely cited sources are not automatically treated as the strongest sources for every market.