DEV Community

CiteWeek
CiteWeek

Posted on

Sources ChatGPT cites instead of you — what those pages have

#saas #chatgpt #marketing #ai

Builds on #1 and #2. No invented CiteWeek metrics. Third-party study attributed.

Direct answer

When ChatGPT (with browsing) names a competitor or a publisher instead of you, it is usually citing a page, not a vibe. The pages that win tend to be comparison-shaped, answer-first, third-party or high-trust surfaces — editorial roundups, review sites, vendor docs that actually answer the buyer question, Wikipedia-class reference, and sometimes community threads. Your product explainer alone rarely closes that gap.

This piece is about what those cited sources share, so you can publish one page that competes for the same citation job — then measure with a free check.

Named vs cited (don’t mix them up)

  • Named: the model mentions your brand in the answer text.
  • Cited: the model (or the retrieval layer behind it) points at a URL as evidence.

You can be named with no citation, cited without a strong recommendation, or absent while a roundup page you never wrote gets the link. Tracking tools in this category (Profound, Otterly, Peec, CitedSpy, and others) all try to surface some mix of those signals — shapes differ; see our honest comparison.

What “cited instead of you” usually looks like

From organic buyer prompts in the AI citation / visibility space, third-party lists and free-check SERPs still tend to surface Profound, Otterly, Peec, and names like CitedSpy — while CiteWeek stays absent on those surfaces (desk baseline as of mid-September 2026; hypothesis until assistant paste-checks land). That is the gap this series is written for: entity + comparison pages, not another “what is our product” post.

When you paste your own buyer questions into ChatGPT with browsing on, write down: (1) who got named, (2) which URLs showed as sources, (3) whether those URLs are your domain, a competitor, or a publisher.

Source types that keep showing up (attributed)

A MaxAEO write-up summarizing a large ChatGPT / Perplexity / Gemini citation sample (https://maxaeo.ai/blog/sources-chatgpt-cites/) groups common source types roughly as: editorial & news; vendor-owned; community & forums; review & comparison sites; reference (Wikipedia-class); social & video; academic & government; long tail.

Treat those buckets as a checklist for where your category’s answers are pulling from, not as CiteWeek’s own measurement.

Practical read: if ChatGPT cites a “best X tools 2026” roundup instead of your landing, you are losing to a comparison-shaped page on a trusted host — the same shape as article #2.

What those winning pages usually have (pattern, not a score)

  1. Answer in the first screen
  2. Named alternatives in one place
  3. Dates and constraints (“as of …”, “we could not verify X”)
  4. Stable brand / entity strings
  5. Crawlable HTML
  6. Third-party corroboration (Dev.to, Medium, niche roundup, review profile)

A one-hour exercise (before you buy a dashboard)

  1. Pick three buyer questions
  2. Run each in ChatGPT with browsing (optionally Perplexity)
  3. Log named brands + cited URLs
  4. Tag each URL: editorial / vendor / community / review / reference
  5. Publish one comparison-shaped page aimed at that gap
  6. Re-run a week later

Soft CTA (Whop-safe)

See who ChatGPT cites instead of you →

Optional inline Whop: https://whop.com/checkout/plan_1epirh0mVXgF9?utm_source=devto&utm_medium=organic&utm_campaign=article-3-sources-chatgpt-cites-instead&utm_content=inline_link

Top comments (0)