The sections below contain the technical evidence behind the narrative above. They document the exact dataset scope, definitions and analytical rules so that the conclusions can be checked rather than accepted as marketing claims.

Linked vs Unlinked Brand Mentions
The complete research article continues below in its original wording.
Abstract
Read section →02Research sectionResearch Question
Read section →03Research sectionWhat changes when a Brand Mention is linked versus unlinked?
Read section →04Research sectionDataset
Read section →05Research sectionGold Dataset v1
Read section →06Research sectionFinding 1 — Unlinked Brand Mentions Were a Substantial Part of Explicit Brand Visibility
Read section →Abstract
Brand Mention analysis is often reduced to whether a link exists.
Research data and evidence

| What this shows | Reading |
|---|---|
| Linked explicit mentions | Brand is named and linked. |
| Unlinked explicit mentions | Brand is named but not linked. |
| Shared property | Both count as Brand Mentions because the brand is explicitly visible. |
| Measurement lesson | Counting only links misses real brand visibility; counting all entity links can overstate it. |
That misses an important distinction.
A brand can be explicitly named and linked.
It can also be explicitly named without any matching outbound link.
In the FunkyMEDIA Gold Dataset v1:
- 416 Explicit Brand Mentions — Linked
- 328 Unlinked Brand Mentions
Both classes contain an explicit brand or entity name.
The difference is whether a matching outbound link accompanies that visible mention.
This study examines why linked and unlinked Brand Mentions should be treated as separate but valid forms of explicit brand visibility.
Research Question
What changes when a Brand Mention is linked versus unlinked?
A linked Brand Mention contains:
- explicit brand naming,
- plus a matching outbound link.
An unlinked Brand Mention contains:
- explicit brand naming,
- without a matching outbound link.
Both tell us:
the user explicitly named the brand.
Only one tells us:
the user also pointed to a specific destination.
This study asks:
- How common are linked and unlinked explicit mentions in the Gold Dataset?
- What information can each type carry?
- What does link presence add?
- What does it not tell us?
- Why should both be measured separately?
Dataset
This study uses the Gold Dataset v1.
Gold Dataset v1
- 1,000 threads
- 4,447 posts
- 744 explicit Brand Mentions
Those explicit mentions were divided into:
- 416 linked
- 328 unlinked
The findings on this page describe the Gold Dataset only.
They should not be generalized to the entire 1.77M inventory as population percentages.
Finding 1 — Unlinked Brand Mentions Were a Substantial Part of Explicit Brand Visibility
Out of 744 explicit Brand Mentions:
- 416 were linked
- 328 were unlinked
That means unlinked mentions were not rare exceptions inside the Gold Dataset.
They formed a substantial share of explicit naming.
This matters because a link-only monitoring system would miss them.
Finding 2 — A Brand Does Not Need a Link to Create Context
Consider:
“I use Brand X for office printing.”
If no link is present, the statement still creates:
Brand X → office printing
Add more detail:
“I use Brand X for office printing because the service is local and running costs are lower.”
Now the same unlinked mention creates:
Brand X → office printing
Brand X → local service
Brand X → running costs
The absence of a link does not remove the semantic relationships.
Finding 3 — A Link Adds Destination Information
A linked Brand Mention contains something an unlinked mention does not:
a destination
That destination can reveal:
- a brand-owned site,
- a product page,
- a service page,
- a location page,
- a specific offer,
- a supporting resource.
This makes links useful.
But destination information should not be confused with semantic depth.
A linked mention can be shallow.
An unlinked mention can be highly detailed.
These are separate dimensions.
Finding 4 — Linked Does Not Mean High-Context
Example:
“Brand X”
linked to the brand website.
This is explicit and linked.
But context depth is minimal.
Compare it with an unlinked statement:
“We have used Brand Y for two years in our small office because the service is local and the devices are reliable.”
The second mention has no link, but far more context.
This is why:
Link Status ≠ Brand Context Depth
Finding 5 — Unlinked Does Not Mean Weak
Unlinked Brand Mentions can appear in:
- recommendations,
- questions,
- comparisons,
- first-hand experiences,
- complaints,
- product discussions,
- local service discussions.
A user does not need to paste a URL to make a brand part of a conversation.
From a language perspective, the explicit brand name is already present.
Finding 6 — Linked and Unlinked Mentions Answer Different Analytical Questions
Linked mentions help answer:
- Where did users point?
- Which destination was associated with the brand?
- Was a product or service URL shared?
- Was a company website referenced?
Unlinked mentions help answer:
- Which brand did users name without linking?
- How often does brand visibility occur in natural text?
- What context surrounds the brand when no URL is present?
- Which products, problems or outcomes are connected to the brand?
The two classes complement each other.
Finding 7 — Link-Only Analysis Can Underestimate Explicit Brand Visibility
If a monitoring system counts only linked mentions, it excludes explicit unlinked naming.
In the Gold Dataset, that would remove:
328 explicit Brand Mentions
from the primary brand-name visibility layer.
That is why linked-only measurement is incomplete for Brand Mention research.
Finding 8 — Counting Every Link as a Mention Creates the Opposite Error
The inverse problem is equally important.
If every brand-owned destination URL is counted as a Brand Mention, explicit brand-name visibility can be overstated.
The same Gold Dataset contained:
378 Linked Entity References
These records linked to an entity but did not explicitly name the brand.
So two opposite errors are possible:
Error 1
Count only links → miss unlinked Brand Mentions.
Error 2
Count every brand-owned link as a mention → overcount explicit Brand Mentions.
The taxonomy solves both problems.
The Three-Class Model
The full taxonomy is:
1. Explicit Brand Mention — Linked
Brand name visible + matching link.
2. Unlinked Brand Mention
Brand name visible + no matching link.
3. Linked Entity Reference
Matching destination link + no explicit brand name.
Only the first two classes count toward the primary Explicit Brand Mention total.
Why This Matters for Brand Monitoring
A more complete Brand Mention system should report:
- linked explicit mentions,
- unlinked explicit mentions,
- linked entity references,
separately.
That creates a clearer picture of:
- explicit naming,
- destination linking,
- entity referencing,
- text-based visibility.
A single number loses that distinction.
Why This Matters for SEO
Links remain a central technical SEO signal.
But Brand Mention analysis asks a different question.
SEO:
Is there a link?
Brand Mention analysis:
Is the brand explicitly named?
Entity-reference analysis:
Is the entity referenced even if the name is absent?
Those dimensions can coexist.
They should not be collapsed.
Why This Matters for Digital PR
Digital PR teams often report:
- placements,
- links,
- mentions,
- linked mentions,
- unlinked mentions.
The FunkyMEDIA taxonomy adds a useful distinction:
Linked Entity Reference
A placement can point to a company without explicitly naming it.
That should be visible in reporting.
It may still carry value.
But it should not be mislabeled as an explicit Brand Mention.
Why This Matters for AI Search Research
The current study does not prove that linked and unlinked mentions affect AI systems differently.
But it creates a clear experimental framework.
Future tests can compare:
- linked explicit Brand Mentions,
- unlinked Brand Mentions,
- Linked Entity References,
- high-context vs low-context versions of each.
Possible questions:
- Does AI associate a brand differently when it is explicitly named?
- Do descriptive entity links create similar semantic associations?
- Does unlinked context correlate with AI representation?
- Does destination specificity matter?
These questions require separate experiments.
Linked vs Unlinked and Brand Context Depth
Link status and context depth should be measured independently.
A linked mention can be:
- BCD 0
- BCD 1
- BCD 5
An unlinked mention can also be:
- BCD 0
- BCD 1
- BCD 5
The presence of a link tells us something about destination behavior.
BCD tells us something about contextual richness.
They are different variables.
Linked vs Unlinked and Recommendation Strength
The same principle applies to Recommendation Strength.
A linked mention can be neutral.
An unlinked mention can be a strong recommendation.
Example:
“I recommend Brand X. We have used them for years.”
No link required.
Recommendation Strength measures endorsement.
Link status measures destination reference.
Again, separate dimensions.
What This Study Does Not Prove
This study does not prove that:
- linked mentions are stronger than unlinked mentions
- unlinked mentions are more natural
- linked mentions are more valuable
- unlinked mentions influence AI systems more
- links are unnecessary
- Linked Entity References have no value
- explicit naming causes rankings
The study establishes something narrower:
Linked and unlinked Brand Mentions are different classes of explicit brand visibility.
Methodology
A record is classified as Explicit Brand Mention — Linked when:
- the brand/entity name is explicitly visible,
- and a matching outbound link is present.
A record is classified as Unlinked Brand Mention when:
- the brand/entity name is explicitly visible,
- and there is no matching outbound link.
A record is classified as Linked Entity Reference when:
- a matching entity destination is linked,
- but the brand name is not explicitly visible.
The taxonomy uses a conservative rule to reduce false inflation of Brand Mention counts.
Main Conclusion
Links and Brand Mentions overlap.
They are not the same thing.
A complete analysis needs to recognize:
- explicit linked mentions,
- explicit unlinked mentions,
- entity references without explicit brand naming.
That leads to the central conclusion:
Brand Visibility Exists With and Without Links
The link is one part of the signal.
The explicit name and the surrounding context are separate parts.
Dataset References
- Gold Dataset v1
- Brand Mention Taxonomy Dataset v1
Author / Publisher
Author: Rafał Cyrański
Publisher: FunkyMEDIA Research Lab
Research hub: funkymedia.space