Documentation
Improving visibility via external publications and mentions
How the command center turns the source-domain map into a plan for external placements: the main recommendation is the "outlet × topic" pair, and the content plan serves as a brief for the editorial team
Open in dashboardExternal publications — the command center's second surface
The External publications tab gathers recommendations for where the brand needs to appear outside its own site. For most commercial queries, AI cites external sources — editorial outlets, catalogs, communities, expert domains — more often than the brand sites themselves. So presence in those sources directly determines whether your brand shows up in an AI response.
- Five classes of external sources: editorial publications and media, catalogs and aggregators, communities (forums, Q&A), reviews and ratings, and authoritative domains (educational, governmental, research). Each class is a separate placement channel with its own rules of entry
- Recommendations are grounded in the Sources page from analytics — the domains AI already cites on your prompts. If AI references vc.ru, Habr, or an industry catalog, those exact outlets land in actions on the external surface: GEO Scout recommends specific outlets relevant to your niche, not "any media"
- The surface switcher at the top of the command center: "Site changes" → "External publications". The choice is remembered — return to the page and you land on the same tab you left. Next to the counter you see how many actions have piled up on each surface
- A working filter under the tab narrows by channel: editorial publications, catalogs, reviews, authoritative mentions. Convenient when handing one batch to PR, catalogs to marketing, and community work to a community manager
Don't treat external publications as an add-on to on-site work. For many niches this channel grows citations faster than on-site changes: AI already trusts these sources — you just need to land in the right material. Always check side by side which surface has more unresolved high-impact actions.
Outlet × topic — the main recommendation
The key artifact of the external surface is an outlet card with a set of recommended topics. The main GEO Scout recommendation isn't "publish an article" — it's "get a mention on vc.ru in this specific topic". The "where × about what" pair is the foundation of the whole external placement plan; the content and content plan come second, as a brief for the editorial team.
- The card header — the domain itself (e.g. vc.ru), the source class (Editorial publication, Catalog, Community…), the domain's total citations for the period, and the number of affected prompts. This is scale context: how much AI favors this source on your topics specifically and whether a placement is worth the investment
- The "N topics" counter — how many semantic topics GEO Scout gathered around the outlet. Topics aren't invented: they emerge from clusters of real prompts where vc.ru is already cited. Example: "Ranking of web and digital agencies in Russia" — 3 prompts, 12 citations. That's a ready brief for the editorial team: which format works on this outlet
- The main recommendation is prioritized by the "outlet × topic" pair: at the top — pairs where AI references the domain most strongly and where you have the biggest gaps. The task isn't "write an article" but "land a brand mention on vc.ru in this specific topic" — through an editor pitch, a guest piece, an expert comment, or inclusion in a roundup
- Status buttons ("Done", "Postpone", "Hide") work at the topic level, not the whole outlet. Mark one topic as closed, postpone another until an editorial window, drop a third as irrelevant — the rest of the outlet's topics stay in play
Read the "prompts × citations" pair as the commercial size of the opportunity. A topic with 3 prompts and 12 citations is worth more than a topic with 1 prompt and 12 citations: in the second, AI goes to the outlet on a single narrow query; in the first, on three different ones — and one mention closes several losing prompts at once.
Evidence base: prompts, coverage, and citation examples
Expanding a topic inside the outlet card reveals the full evidence base for the recommendation. You see which prompts are affected, your current coverage on each, and which materials AI already references. It's both the priority justification and a reference for an editor or PR specialist.
- "Affected prompts" — the list of prompts where this outlet is already cited and where a brand mention needs to be added. Next to each prompt is the current brand coverage ("0%", "50%", "100%"). 0% means the brand isn't appearing in AI responses on this prompt — the largest gap and the highest priority
- "Citation examples" — real URLs on the outlet that AI already picks as a source on these prompts. For example, `vc.ru/marketing/…top-30-digital-agentstv-v-rossii`. It's a live reference for the editorial team: this is the format, length, and structure of materials that actually land in AI responses
- Links open in new tabs — read the articles immediately, see the authors, figure out which editor to pitch. On catalogs and review sites, the same links show the card structure and the type of content that gets evaluated there
- The numbers under each topic (prompts × citations) help you choose which topic on the outlet to start with: with many prompts and low coverage, one good placement closes several gaps at once; with few prompts and 50–100% coverage, push to the next iteration
0% current coverage means "we're not in the answer", but AI already considers the outlet authoritative. That's the cheapest situation to close: the source is trusted, you just need to land in the right material. 50–100% coverage means the outlet already knows you — quick gains there are exhausted, top up via other topics and domains.
Content plan and the content itself — as a brief for the editorial team
The "Generate content plan" button on a topic card is a supporting tool, not the main recommendation. The main action is still to land a brand mention on the recommended outlet in this topic. The content plan and the ready content help articulate how the material should be put together so AI recognizes it as relevant and includes it in answers.
- The content plan is built from monitoring data: structure, key entities, facts, quotes, and the prompts the text should target. It isn't a template-article spec but a brief for an editor or PR specialist — what must be mentioned and which meanings to surface so the material works for citability
- Unlike pillar pages on your own site, here you don't publish the text yourself. The end point is an agreement with the outlet's editorial team, a guest publication, an expert comment in an upcoming roundup, an inclusion in a catalog, or a community answer. The content plan and the content itself act as ammunition in the pitch, not the final artifact
- Action statuses are the standard ones: "Done" (the placement happened), "Postpone" (waiting for an editorial window or author reply), "Hide" (outlet irrelevant, conflict of interest, too expensive). History is preserved — return a month later and reactivate the topic
- The loop closes on the next monitoring iteration. If AI starts citing a fresh piece on the recommended outlet that mentions the brand, the signal is confirmed: domain Citation Share grows, and the Sources map reflects the new citability across the very prompts the placement was made for
Treat the generated content plan as a brief, not a finished text. The outlet knows its formats, authors, and editorial policy better than you — a short pitch with the key meanings from the plan usually works better than a finished article sent in whole. The goal is a brand mention in the right topic, not the publication of your text.
Tracking the impact of publications: closing the loop
Once the material is out — on an external outlet or your own site — add its URL to Marketing campaigns to track how AI reacts: how many times it landed in AI responses, on which prompts, and what citation lift it produced. This is exactly where the loop closes: recommendation → publication → measurable impact.
- Marketing campaigns is a separate section for URL tracking. The "+ Add URL" button accepts both external placements (an article on vc.ru, a catalog card, a community thread) and pages on your own site — pillar long-reads, child articles, new service cards. A campaign groups placements within a single release, PR wave, or quarterly content plan
- Three KPI cards at the top of the campaign: "Citations" — how many times the URL landed in AI answers; "Citation rate" — what share of AI answers on the tracked prompts contains this URL; "Prompts" — in how many distinct queries AI used the page. An instant answer to "did the placement work"
- The "URL sources" tab shows a table of tracked links broken out by domain: citations, response share, and prompt count per URL. Expand a domain to see the specific page (`vc.ru/marketing/…top-30-digital-agentstv-v-rossii-2026`) — which exact material is bringing citations and which is sitting idle
- The "Prompts" tab shows in which exact queries AI cites the placement. Useful for cross-checking with command-center priorities: if the material closes high-priority prompt gaps, the loop is closed; if it lands on irrelevant ones, that's a signal of a content or topic mismatch
This is the metric that makes GEO work measurable: a recommendation from the command center → publication → URL in a campaign → citations in the next monitoring cycle. If after 2–3 cycles a URL doesn't pick up citations, that's the signal to rework the material or try another outlet — and only confirmed placements with real impact move into "Done", not the mere fact of publication.