Owned vs Earned Mentions

Search and AI assistants have split the attention economy. Your SEO dashboards show click-throughs and rankings, but they rarely tell you whether a generative model is quoting your article, citing a review, or repeating a Wikipedia summary. That gap matters now because more buying journeys start with a conversational prompt. When an AI names a competitor instead of you, the loss of consideration can happen before your site ever loads.

Understanding the Distinction

Owned mentions are direct citations of content you control. Think of a model quoting a how-to post from your blog or linking to product documentation. You decide the messaging, the depth, the examples, and when it gets updated. Owned mentions are the place to push technical detail, playbooks, and original data that make a model’s answer point back to you.

Earned mentions come from third-party sources: reviews, news articles, user forums, and public databases. A model may say, "According to customer reviews, X scores highest for ease of use," and that references earned signals. Those mentions are often stronger trust signals because they’re independent, but you don't control timing or framing. Earned mentions tend to persist in training corpora and public knowledge graphs, so their effect can be long lived.

Building Owned Mention Assets

Owned assets are the quickest way to drive direct citations. Prioritize content that answers one of three buyer needs: factual definitions, procedural steps, and comparative analysis. Models prefer concise, authoritative sources that are well-structured and include clear attribution. A product page that only lists features will rarely be cited; a short technical explainer with a firm conclusion and examples will.

• Make pages that answer single questions, with a clear headline and a canonical URL.
• Include dated references and versioning on technical topics so models can pick the freshest signal.
• Add short, copyable snippets such as TL;DR boxes or plain-text summaries that are easy for parsers to extract.
• Expose structured data where appropriate, like FAQ schema and product schema, but focus on clean prose first.

And don’t forget distribution. Syndicate white papers to platforms known to be scraped by models, and gate the right content only if you want it excluded from quick citation.

Cultivating Earned Mentions

Earned mentions are slower and messier, but they carry independent credibility. PR still works, and targeted review campaigns move the needle. Get reviewers to speak to specifics, not generic praise, because detail is what models copy into answers. A one-line endorsement is weaker than a paragraph that contrasts your product on price, support, and integration.

Target the sources that AI systems ingest or that feed knowledge graphs: reputable trade sites, major review platforms, and community forums. Encourage contributors to use your exact brand name and product names consistently. Monitor Wikipedia and public datasets, and correct factual errors with proper citation and talk-page communication where relevant. When you earn coverage, capture screenshots and crawl copies so you can trace where an AI mention likely came from.

Measuring Both Types

Track owned and earned separately because they behave differently. For owned mentions, measure citation frequency, citation depth, and top queries that led to the citation. For earned mentions, count appearances across third-party sites, sentiment of those mentions, and presence in authoritative knowledge sources. Combine those with downstream signals such as changes in branded query volume, conversions attributed to assisted channels, and shifts in SERP features.

Use a mix of methods. Automated scrapes and link monitoring will catch direct links and text matches from your assets. API queries against major answer engines and sampling of model responses for representative prompts will show what models actually say. Manual audits remain useful for earned signals; annotate which outlets are more likely to influence model answers, and weight them accordingly.

• Owned metrics: citation count, citations per asset, time-to-citation after publish.
• Earned metrics: mentions by domain authority, sentiment, inclusion in knowledge panels or encyclopedic sources.
• Combined view: share of mentions by type, conversion lift after major earned placement.

Report on both monthly. A spike in owned citations after a documentation release suggests content is being picked up quickly. A steady rise in earned mentions from reviews indicates growing social proof. Both feed visibility in generative answers, but you manage them in different ways.