Module 04 · Citations

Citation intelligence for AI search.

Every source AI engines cite — classified, scored, and tagged as brand-supporting, competitor-supporting or a gap you can close.

11 source types · 4 engines · live graph
Citation graph · 30d
resolving
g2.comPeer reviewtechcrunch.comEarnedreddit.comUGCycombinator.comIndustryCitationsSIS 84
Earned
Peer
UGC
The citation layer

Sources, scored.

Source types
0
Deterministic classification across every cited domain — no manual tagging required.
Median citations / week
0
Across a Pro-plan brand's tracked prompt set.
Top SIS
0
Source Influence Score — what "leader" looks like in a competitive category.
11 source types

Every cited domain, deterministically classified.

So the same domain gets the same classification on every rescan — no drift, no analyst disagreement.

01
Earned media

Editorial coverage on TechCrunch, The Verge, industry trades.

02
Peer review

G2, Capterra, TrustRadius and category-specific review platforms.

03
Affiliate

Wirecutter-style affiliate listicles and best-of round-ups.

04
Competitor

Competitor-owned pages cited as authoritative.

05
Social

X, LinkedIn and other social platforms surfaced by AI engines.

06
Brand

Your own pages — and the pages that link to them.

07
Review

Long-form expert reviews and tear-downs.

08
Directory

Crunchbase, Built In, Product Hunt and similar listings.

09
UGC

Reddit, HackerNews, Quora and other community surfaces.

10
Industry

Analyst reports, association publications, industry indexes.

11
Partner

Integration directories and partnership announcement pages.

Brand side
+

Sources that consistently cite, recommend or rank you. Defend and expand.

Competitor side

Sources that cite competitors at your expense. Pursue placement here first.

Neutral side
·

Sources cited in your category but not favoring either side. Opportunity territory.

Engines
4
ChatGPT, Perplexity, Gemini and Google AI Overviews out of the box.
Sources tracked
5k+
Median unique domains cited across a 30-day prompt set.
Attribution
URL-level
Every citation is resolved to the exact page and section.
Sentiment
±3-tier
Positive, neutral or negative context around each citation.
Citation pipeline
Step 01
Sample

Prompt runs across engines with n=10 samples for stability.

Step 02
Extract

URLs, domains and anchor context pulled from every answer.

Step 03
Classify

Source type, sentiment and rank position per engine.

Step 04
Attribute

Roll up to brand, domain and page-level share of voice.

Do
  • Track citations weekly — engine indexes shift constantly.
  • Segment citations by engine — Perplexity and ChatGPT weight sources differently.
  • Reverse-engineer top-cited pages into your own content briefs.
  • Flag negative-sentiment citations for a PR response.
Don't
  • Don't celebrate raw citation counts — position and sentiment matter more.
  • Don't ignore third-party listicles that cite competitors but not you.
  • Don't treat a single sample as truth — always look at n≥5 runs.
  • Don't optimise for one engine at the expense of the other three.

Keep reading

FAQ

Questions about citations.

How are sources classified?+
A deterministic classifier maps every domain to one of 11 types — earned, peer, affiliate, competitor, social, brand, review, directory, UGC, industry, partner.
What's the Source Influence Score?+
A composite of frequency × prompts reached × engines reached × recommendation weight × side boost (brand / competitor / neutral).
Can I tag custom sources?+
Yes — override classification per domain, and the override persists across rescans.
Do you track snippet-level citations?+
Yes. Every citation captures URL, domain, snippet, surrounding context and sentiment.
Try free

Map your citation landscape.

5-day trial for $5. Every cited source, classified and scored, the day you sign up.