NeverRanked · Directory playbooks

What AI engines extract from each third-party source.

Per-directory playbooks for the 12 surfaces that surface as recurring third-party hits across the 5 measured categories. For each: which fields matter, what extracts into AI citations, what is decoration, what cadence to maintain. Useful when your per-query playbook says "third-party content dominates this query."

Use this when: the AI-citation data for one of your queries shows third-party sources dominate (e.g., 60% of the cites for a query go to a directory or association, not to firm websites). Your firm cannot win that query by improving the firm's own site alone. You also need to be properly represented on the third-party surfaces AI cites. This page is the field-by-field optimization guide for each one.
What is measured vs what is inferred. The citation counts and cohort-wide patterns named on this page (e.g., "Microsoft Copilot 0-2% own-share across 5 categories", "HSCPA cited 130 times in 3 runs", "Clutch top B2B services source") are direct findings from NeverRanked's measurement runs. The field-by-field optimization advice for each source (e.g., "respond to every review within 7 days", "post weekly") is a combination of: (a) NeverRanked's observation of which fields AI engines extract, (b) each platform's public documentation, and (c) generally-accepted directory-optimization best-practice. Field-level advice is informed prior. Your monthly measurement will confirm or correct what actually moves citations for your specific category.
Where the line is. NeverRanked publishes the playbooks as a spec. The customer's team (or the customer's agency) claims the profile, populates the fields, responds to the reviews, and maintains the cadence. NeverRanked does not have logins to your Google Business Profile, your Bing Business Profile, your LinkedIn page, your Clutch profile, or any other third-party platform on your behalf. The boundary is structural: we measure, we hand off a spec, the customer or agency executes. That separation is the engagement.

Sources covered

Google Business Profile (GBP)

For: every category. Universal. The single highest-leverage third-party surface for AI citation.
Primary AI engine pull
Google AI Overviews (primary), Perplexity (secondary), Gemini grounded (secondary)
Why it matters
Google AI Overviews grounds local queries in Google's own data stack. GBP is the canonical input. Strong GBP raises citation share across multiple engines, not just Google's own.
Effort to set up
2-4 hours initial. 30 min/month to maintain.
Cost
Free.
What matters (AI extracts these)
  • Business name. Must match the firm's primary brand name exactly. Mismatches with the website penalize both.
  • Primary and secondary categories. Pick the most specific category available. AI engines use this for category-shaped queries.
  • Service list. Add every distinct service. AI cross-references with website Service schema.
  • Hours. Specific open/close times per day, special hours for holidays. AI surfaces "open now" facts directly from this.
  • Reviews and responses. Volume and recency matter. Responding to reviews (every one, within 7 days) signals active maintenance.
  • Photos. Refresh quarterly. AI engines weight active profiles higher than dormant ones.
  • Q&A. Seed 5-8 buyer-shaped questions with answers. Public Q&A from buyers, answer within 48 hours.
  • Posts. Weekly cadence. Short, specific. AI uses these as freshness signal.
What is decoration (do not over-invest)
  • Cover photo aesthetics beyond "clear and professional"
  • Holiday-special posts and seasonal banners
  • The "products" tab unless you sell discrete inventory
Cadence Weekly: post one short update. Monthly: respond to any unanswered reviews and new Q&A. Quarterly: refresh photos and service list. Annually: full audit against the AI-citation playbook.

Bing Business Profile (Microsoft Copilot lever)

For: every category. The single highest-leverage move for Microsoft Copilot citation, which sits at 0-2% own-site share cohort-wide across all 5 measured categories.
Primary AI engine pull
Microsoft Copilot (almost exclusively, since Copilot answers using Bing organic + Bing Business data)
Why it matters
Across every category NeverRanked has measured, Microsoft Copilot cites firm-owned websites at 0-2% rate. The first-mover advantage on Bing organic + Bing Business is intact because most competitors are not present.
Effort to set up
1-2 hours. The platform is less polished than GBP and most firms skip it. That is the opportunity.
Cost
Free.
What matters
  • Business name + address + phone. The NAP triplet. Must match GBP and website exactly.
  • Category. Same as GBP, mapped to Bing's category taxonomy.
  • Hours. Microsoft Copilot surfaces "open now" facts.
  • Verified status. Verify the profile (postcard or phone). Unverified profiles are deprioritized.
  • Website URL. Use the canonical domain, not a tracking parameter.
  • Description. 200-300 words. Plain language. Mirror the firm's website meta description.
What is decoration
  • Photos beyond a clear primary image
  • Bing-specific posts (these get cited rarely)
  • The "menu" or "appointment" extensions for non-applicable categories
Cadence Set up once, audit annually. Bing Business changes less than GBP. Over-tending is wasted effort.

LinkedIn Company Page

For: B2B-heavy categories (CPA, law, wealth, agencies). Less critical for B2C local service.
Primary AI engine pull
ChatGPT search, Gemini grounded, Google AI Overviews. Secondary for Perplexity
Why it matters
LinkedIn shows up as a recurring third-party citation across B2B-heavy categories. AI engines use it to confirm entity identity, named professionals, and B2B credibility signals.
Effort to set up
3-5 hours. Ongoing 1-2 hours/month for posts.
Cost
Free company page. LinkedIn Premium not required.
What matters
  • Company name. Exact match to brand. Mismatches with website confuse AI's entity resolution.
  • About section. 300-500 words. Plain language. Lead with what the firm does, then named services, then named buyer types.
  • Specialties. List 5-10 specific service or subject areas. AI cross-references with Service and Person schema on the firm's website.
  • Employees. Connected named partners with completed profiles. Especially important for Person-driven categories (law, wealth, CPA).
  • Posts. Weekly cadence. Even short text-only posts maintain freshness signal.
  • Headquarters address. Match LocalBusiness schema on the firm's website.
What is decoration
  • Logo aesthetics beyond clarity
  • LinkedIn-native video unless the firm already produces video
  • The "products" section for service businesses
Cadence Weekly: one substantive post. Monthly: review employee profile linkage. Quarterly: refresh About section if positioning has shifted.

Yelp

For: consumer-facing categories (med spa, dental, restaurants, retail). Less critical for B2B professional services.
Primary AI engine pull
Perplexity, ChatGPT search, Google AI Overviews
Why it matters
For consumer-facing categories, Yelp is a recurring third-party citation source. AI engines use it for review-driven queries ("best", "top rated", "reviews of").
Effort to set up
1-2 hours. Ongoing minimal except review response.
Cost
Free profile. Paid advertising not required for AEO purposes.
What matters
  • Claimed business status. Unclaimed profiles get deprioritized by AI engines.
  • Categories. Specific. Match what GBP uses.
  • Service list. Specific service variants where applicable.
  • Review responses. Respond to every review within 7 days. AI engines treat response rate as a freshness signal.
  • Photos. 10+ high-quality photos minimum.
  • Hours. Match GBP and Bing.
What is decoration
  • Yelp ads (do not improve AI citation share)
  • Yelp Deals unless they fit the business model
  • Custom Yelp page banners
Cadence Weekly: review response. Quarterly: photo refresh. Annual: full profile audit.

Better Business Bureau (BBB)

For: trust-sensitive categories (CPA, law, wealth, home services). Showed up at 30 cites across 3 runs in CPA Hawaii data.
Primary AI engine pull
Google AI Overviews, Perplexity, ChatGPT search
Why it matters
AI engines cite BBB profiles for trust-shaped queries (especially for high-consideration purchases). The accredited badge signal carries weight in AI's source ranking.
Effort to set up
2-3 hours for free profile. Accreditation involves a fee and a review process.
Cost
Free unaccredited profile. Accreditation fees scale by firm size (typically $500-$5,000/year).
What matters
  • Accredited status. The badge appears in AI citations and lifts the source's apparent authority.
  • Complaint resolution rate. Respond to and resolve every complaint within 30 days.
  • Business description. 100-200 words. Plain language match to website.
  • Contact info NAP. Must match GBP, Bing, website.
What is decoration
  • BBB-branded marketing materials beyond the badge on your own website
  • Posts and updates on the BBB profile itself
Cadence Monthly: complaint and inquiry review. Annual: accreditation renewal and profile audit.

Clutch.co

For: B2B services (CPA, agencies, IT services, consulting). Surfaced at 101 cites across 3 runs in CPA Hawaii data, making it the top review-directory source for that category.
Primary AI engine pull
ChatGPT search, Perplexity, Gemini grounded
Why it matters
Clutch is one of the most cited B2B review directories for professional-services queries. AI engines treat the editorial-leaderboard structure as a trust signal.
Effort to set up
3-5 hours for profile + at least 3 verified client reviews to be eligible for category leaderboards.
Cost
Free profile. Paid sponsored placement available but not necessary for AEO.
What matters
  • Verified client reviews. Minimum 3 to unlock the leaderboard. AI engines cite Clutch leaderboard pages more than firm profile pages.
  • Service offerings. Specific services tagged correctly.
  • Industry focus. Pick 2-4 specific industries; "all industries" reads as weak signal.
  • Client list. Named clients (with their permission) carry weight.
  • Case studies. 2-5 written case studies with measurable outcomes.
  • Geographic location. Match website LocalBusiness schema.
What is decoration
  • Banner and visual styling of the profile
  • Clutch-paid badges beyond the legitimate "Top X" recognition
  • The community-engagement features
Cadence Quarterly: solicit 1-2 new client reviews to keep the review velocity active. Annual: refresh case studies and service list.

Expertise.com

For: multi-category service businesses. Showed up at 35 cites across 3 runs in CPA Hawaii data.
Primary AI engine pull
Perplexity, ChatGPT search, Google AI Overviews
Why it matters
Expertise publishes editorial "best of" lists per category per city. AI engines cite these list pages frequently for "best in [city]" queries.
Effort to get listed
Editorial selection, not self-service. Expertise contacts firms. Firms can also pitch with strong portfolio + reviews.
Cost
Free to be listed if selected. Paid placement also available but not required.
What matters
  • Being on the list at all. Inclusion is the primary lift. Placement order is secondary.
  • Reviews and reputation across other surfaces. Expertise's editorial team uses GBP, Yelp, and BBB reviews as selection inputs.
  • Years in business. Established firms are favored in Expertise's editorial logic.
  • Specialty clarity. Firms with one clear specialty get listed more often than generalists.
What is decoration
  • Pitching aggressively for inclusion (this is editorial, not paid)
  • Paid placement (the editorial selection is what drives AI citation weight)
Cadence No active maintenance. Annual: check whether you have been added to any new Expertise lists for your category and city.

Wikipedia

For: established firms with documented public history. The most defensible AEO asset but hardest to earn.
Primary AI engine pull
Claude and Gemma (heavy weight in training data), Perplexity, Microsoft Copilot, Google AI Overviews
Why it matters
Wikipedia presence compounds in training-data engines over years. Hawaii consumer banks (which have decades of Wikipedia presence) see 88% Claude own-share. Categories without Wikipedia presence (like Hawaii CPAs) see 1-2%.
Effort to earn
Variable. Notability requirements must be met. Independent editors typically write the article. Firms cannot self-publish.
Cost
Free to be the subject of an article. Paying editors to write or maintain articles violates Wikipedia's terms.
What matters
  • Notability evidence. Independent press coverage, industry awards, milestone events. Wikipedia editors evaluate this before creating articles.
  • Reliable sources. Articles need multiple independent secondary sources to support claims.
  • Sustained accuracy. Once an article exists, the firm cannot directly edit. It can submit edit requests via the article's Talk page.
  • Wikidata entry. Adjacent to the article. Some AI engines read Wikidata directly. Ensure key facts match the article.
What is decoration (and dangerous)
  • Self-editing or paying editors to write or modify the article (violates Wikipedia terms. Gets articles deleted)
  • Adding promotional language (editors revert quickly)
  • Trying to remove unflattering but sourced content
Cadence Quarterly: review article for factual accuracy. If incorrect, post a polite edit request to the Talk page citing reliable sources. Do not edit directly if you are affiliated with the firm.

Healthgrades

For: medical and dental categories specifically. Surfaces in dental data alongside Yelp and Google.
Primary AI engine pull
Perplexity, ChatGPT search, Google AI Overviews
Why it matters
For medical and dental queries, Healthgrades is a recurring third-party citation. Practitioner profiles with strong review velocity get cited for "best [specialty] in [city]" queries.
Effort to set up
2-3 hours for profile claim and population. Practitioner-by-practitioner if the practice has multiple providers.
Cost
Free basic profile. Sponsored placement available but not required.
What matters
  • Practitioner profile completeness. Photo, education, board certifications, specialties, insurance accepted.
  • Patient reviews. Actively solicit. Respond to every one.
  • Office address and hours. Match GBP exactly.
  • Languages spoken. Healthgrades surfaces this for specific queries.
What is decoration
  • Sponsored placement (does not lift AI citation share)
  • The "ask a question" feature unless you commit to answering
Cadence Monthly: review response. Quarterly: profile refresh.

Chambers / BestLawyers / Justia (legal-specific)

For: law firms only. Chambers and BestLawyers are editorial rankings. Justia is a self-service legal directory.
Primary AI engine pull
All web-searching engines for legal queries. Chambers + BestLawFirms together accounted for 339 cites across 3 runs in Hawaii law-firm data.
Why it matters
Editorial legal directories are the highest-cited third-party sources for legal queries. AI engines treat the editorial peer-rating process as a strong trust signal.
Effort to engage
Chambers and BestLawyers involve peer-rating surveys that take 10-20 hours of partner time per cycle. Justia is self-service.
Cost
Chambers + BestLawyers free to be ranked (the survey is the cost). Marketing materials and badges cost extra. Justia free basic, paid premium.
What matters
  • Chambers / BestLawyers nomination cycle participation. Submit nominations every annual cycle.
  • Attorney bio completeness on each platform. Education, bar admissions, practice areas, notable cases, publications.
  • Practice area specificity. Specialists get cited more than generalists.
  • Justia profile with practice area content. The free profile is sufficient. Paid premium adds marginal value for AEO.
What is decoration
  • Buying Chambers / BestLawyers branded marketing materials (the ranking is the asset)
  • Justia paid premium tiers for AEO purposes specifically
  • Aggressive nomination campaigns (peer rating is editorial, not gameable)
Cadence Annual: Chambers and BestLawyers nomination cycles. Quarterly: Justia profile review and case-result updates.

State CPA societies (HSCPA and equivalents)

For: CPA firms specifically. HSCPA (Hawaii Society of CPAs) surfaced at 130 cites across 3 runs in Hawaii CPA data, the single most-cited third-party source for that category.
Primary AI engine pull
All web-searching engines for CPA queries. Particularly strong in Perplexity and Google AI Overviews.
Why it matters
State CPA society directories are the editorial trust layer AI engines defer to for CPA queries. A firm not represented in the state society's public directory is invisible across 6%+ of all category citations.
Effort to engage
Active society membership required. Profile completion takes 1-2 hours.
Cost
Membership dues scale with firm size. Typically $300-$800/year per CPA member.
What matters
  • Active member status. Lapsed members disappear from the public directory.
  • Firm profile completeness. Services, specialties, named partners, location.
  • Continuing education credit reporting. Society directories sometimes display CE compliance status. Visible compliance is a trust signal.
  • Participation in society events. Speaking, writing for the newsletter, committee work compounds into editorial mentions over time.
What is decoration
  • Society-branded marketing materials beyond confirming membership
  • Society-paid advertising tiers
Cadence Annual: membership renewal and profile audit. Quarterly: consider one society event participation (speaking or writing).

SmartAsset / Plannersearch / Unbiased (wealth lead-gen)

For: wealth management firms specifically. These surfaced as the lead-gen middleman pattern in Hawaii wealth-management data: AI cites them MORE than individual firms for "best advisor" queries.
Primary AI engine pull
Perplexity, ChatGPT search, Gemini grounded, Google AI Overviews
Why it matters
AI engines preferentially cite these aggregators for wealth-management queries because they aggregate cross-firm information AI can compare. A firm not represented here loses a significant share of AI citation visibility.
Effort to engage
1-2 hours per platform for free listing. SmartAsset and Plannersearch sell paid lead-gen subscriptions ($500-$3,000/month) which include enhanced visibility.
Cost
Free basic listings. Paid lead-gen is its own ROI calculation independent of AEO.
What matters
  • Listed status at all. Inclusion is the primary AEO lift.
  • Specialty fields. Match the firm's actual specialties. Aggregators surface niche matches.
  • Geographic service area. Specific cities/regions.
  • AUM range. Honest representation. AI uses this for "for high-net-worth" queries.
  • Credentials. CFP, CFA, CPA cross-credentials listed.
  • Fee structure transparency. Fee-only vs commission-based vs fee-based clearly stated.
What is decoration
  • Aggressive paid lead-gen tier purchase for AEO purposes alone
  • Aggregator-branded marketing materials
Cadence Quarterly: profile review across all three. Annual: full audit and updates.

The pattern that holds across all sources

  1. NAP consistency. Name, Address, Phone must match exactly across every directory + your website. AI engines cross-reference. Mismatches downweight both signals.
  2. Active maintenance signals trust. AI engines deprioritize stale profiles. Even minimal recurring updates lift citation share.
  3. Review velocity matters more than absolute count. A profile with 20 reviews and 2 new ones this month outperforms a profile with 200 reviews and none this year.
  4. Editorial inclusion beats paid placement for AEO. AI engines weight earned visibility over paid. Spend your time on the editorial path even when paid options exist.
  5. Match what your per-query playbook flags. A directory that does not surface in your category's AI-citation data is not worth optimizing for AEO. Spend time on the directories AI actually cites for your queries.

What happens next

  1. Identify which directories matter for your category from your NeverRanked per-query playbook (the "third-party content dominates" notes).
  2. Audit each one against the playbook above. Fix what is broken or missing.
  3. Establish the cadence for ongoing maintenance.
  4. Wait 4 to 8 weeks for AI engines to re-index after material updates.
  5. Verify in your monthly NeverRanked measurement that the directory-driven queries now cite your firm more often.

Schema library · Page templates · How NeverRanked measures · Return to NeverRanked

NeverRanked · Honolulu · Lance Roylo · Lance@hi.neverranked.com