Your Competitor Gets Named on Pages Neither of You Own
Most competitor research stops at the property line. You pull their keywords, read their service pages, check their review count, and build a list of things to copy. That work has a ceiling, because the pages that decide who gets picked increasingly are not owned by either of you.
Search Engine Journal reported on September 30, 2026 that Wikipedia, podcast pages, and Reddit threads now drive a measurable share of AI citations. Answer engines are not just reading your site. They are reading what other people wrote about you, and about the company down the road. If your competitor appears on twelve third-party pages and you appear on two, you are losing a race you were never auditing.
Run the Buyer's Five Queries, Not Yours
Start with the searches a real buyer runs before they trust anyone. Not head terms. The messy, comparative, skeptical ones:
- "best [service] in [city]" to surface roundups and directory pages
- "[competitor name] reviews" to surface forums, threads, and complaint pages
- "[service] near me recommendations reddit" to surface peer discussion
- "who to hire for [specific problem] [city]" to surface local news, association listings, and community groups
- "[service] cost [city]" to surface pricing content written by third parties
Run each one in Google, then run the same five in an AI assistant and note which sources it cites. Open every result on page one plus the cited sources, and record every page that names a competitor and is owned by neither of you. You are looking for Reddit and forum threads, local news features, chamber and trade association directories, "best of" roundups from regional publishers, podcast episode pages with show notes, supplier and manufacturer partner lists, and permit or licensing databases that render as public pages.
The list is usually shorter than owners expect and more reachable than they fear. Ten to twenty pages is a complete picture for most local businesses. For a multi-location operation, run the same five queries per market and expect the pattern to vary by city, which is exactly the finding that tells you where your field teams have relationships and where they do not.
Build the Mention Map
One spreadsheet, one row per page. Columns:
- Source URL and source type (forum, news, directory, roundup, podcast, partner list)
- Crawlable and dated, meaning the page is publicly indexable and carries a visible publish or update date
- Names competitor (yes/no, and how they are described)
- Names you (yes/no)
- Claim corroborated, meaning the specific fact the page confirms: licensed, serves this zip code, handles this job type, priced in this range, open since this year
- Reachability tier (covered next section)
- Owner and next action date
The crawlable and dated columns matter more than they look. A page locked behind a login, rendered entirely in JavaScript, or undated is weaker corroboration for both search engines and AI systems. A dated, crawlable page from a recognizable local publisher does more work than five directory listings nobody reads.
The claim column is where the strategy lives. You are not counting mentions. You are mapping which facts about your competitor have independent backing and which facts about you have none. If four third-party pages confirm that your competitor handles emergency work after hours and zero pages confirm the same about you, that is the gap. The fix is not a new page on your site claiming it louder.
Score Every Gap by How Reachable It Is
Sort the map into four tiers and give each a different action.
Tier 1: Correctable listings
Directories, association pages, supplier partner lists, and chamber profiles where your record is missing, outdated, or wrong. These are forms and emails. Fix the name, address, phone, service scope, licensing number, and coverage area so they match your site exactly. Same-week work.
Tier 2: Reply-worthy threads
Reddit, Facebook groups, Nextdoor, and trade forums where someone asked a question your competitor got named in. Reply as yourself, with your business identity visible, answering the actual question with useful specifics. If the thread is older than a year, leave it alone and log the question instead.
Tier 3: Roundups with a stated inclusion process
Many regional "best of" lists publish criteria: years in business, license verification, review thresholds, nomination windows. Read the criteria, confirm you meet them, and submit through the stated process. Track the nomination date in the map.
Tier 4: Earned coverage
Local news, podcasts, and industry features that require an actual story. A hiring milestone, an unusual job, a seasonal advisory, a partnership with a known local institution. These take the longest and return the most, because they are dated, crawlable, and written by someone with editorial standing.
Supply Facts, Not Adjectives
Every outreach touch carries the same payload. Third-party corroboration only helps when the third party repeats facts your site already states in the same form. SEOGOD horizon signals point the same direction: AI answer engines select businesses using verified facts that hold up outside the site, and local discovery is shifting from keyword matching to proof.
Build a one-page fact sheet and attach it to every submission, reply, and pitch:
- Legal business name exactly as it appears on your site and license
- License and certification numbers with the issuing body
- Service scope in concrete job types, not category labels
- Coverage area as named cities or zip codes, not a radius claim
- Pricing ranges for common jobs, with what moves a job to the high end
- Year founded, team size, and insurance status
- Response window for emergency or urgent work
When a directory, a roundup, and a podcast episode page all state the same license number and coverage area as your service pages, you have built an entity that holds together under verification. When each source says something slightly different, you have built noise.
Two things stay off the table permanently. Do not astroturf forums with unmarked accounts that recommend your own business. Reddit moderators catch it, and so do the buyers who read the account history. Do not buy roundup placements from publishers selling spots as advertising, because paid inclusion lists lose the editorial standing that made them worth appearing on.
Mention Map Checklist
- Five buyer queries run in both Google and an AI assistant, with cited sources recorded
- Every third-party page naming a competitor logged with URL and source type
- Crawlable and dated status confirmed for each row
- Specific corroborated claim written out, not a yes/no mention count
- Your presence or absence marked on every row
- Reachability tier assigned to all four tiers, with no row left unsorted
- Fact sheet built and matched line by line against your own site copy
- Owner and next action date assigned to every tier 1 and tier 2 row
- Unanswered buyer questions pulled from threads into a separate tab
- Re-run date set 30 days out
What to Fix First
Work in this order, because the early steps make the later ones credible.
- Reconcile your own facts. If your site lists one coverage area and your Google Business Profile lists another, fix that before you ask anyone else to publish anything. Mismatched facts corroborated widely are worse than no corroboration.
- Clear every tier 1 listing. Correctable records are free, fast, and they establish the baseline facts every other source will echo.
- Answer two tier 2 threads per week. Pick the ones where the asker has not been helped yet. Volume is not the goal.
- Submit to roundups during their stated window. Calendar the nomination dates from the map.
- Convert unanswered questions into briefs. Any question in those threads that your site does not answer becomes an owned page with original detail, not a rewrite of your competitor's page.
That last step is the one that compounds. SEOGOD's editorial direction for this cycle is explicit about it: competitor evidence should become unique buyer questions, missing proof, and internal links rather than duplicate pages. A mention map surfaces real questions from real buyers, which is a better brief source than any keyword tool.
The 30-Day Loop
Set a recurring task. On the same day each month, re-run the five queries, add new rows for any page that appeared since the last pass, and check whether your tier 1 corrections actually published. Move completed rows to a closed tab so the active map stays short.
Three things come out of every loop: new mentions earned, listings that drifted back out of sync, and fresh buyer questions headed for the content queue. Inside SEOGOD, those questions land as briefs weighted by conversion intent, so off-site findings feed the publishing queue instead of sitting in a spreadsheet. If you want the baseline before your first loop, start with a free audit and check which facts your own pages state clearly enough to be corroborated at all.
Ready to Stop Guessing?
Run a SEOGOD audit on your domain and see the next proof-backed SEO opportunities.
Start Free Audit