← Back to the Bob Manthy hub

1. The thesis: local leads, not traffic

Bob doesn't sell a product you can ship. He runs a room. People sit in that room for ten weekly three-hour sessions, so they have to be able to get to it and back on a weeknight. That single fact governs the entire SEO plan.

The catchment is two overlapping circles, both about a 60-minute drive:

• Niwot circle (Boulder area cohort): centered near 40.10, -105.17. Covers Boulder County and the north Denver metro edge: Boulder, Longmont, Lafayette, Louisville, Superior, Erie, Broomfield, Lyons, Niwot, Gunbarrel, plus the Carbon Valley towns (Firestone, Frederick, Dacono, Mead) and Berthoud.
• Cherry Creek circle (Denver cohort): central Denver. Covers the Denver metro: Denver proper, Cherry Creek, Glendale, Wash Park, Westminster, Thornton, Northglenn, Arvada, Wheat Ridge, Golden, Englewood, Littleton, Centennial, Highlands Ranch, Lakewood.

People might drive further for a graduate potluck or a hike. They will not drive 90 minutes each way, ten times, for class. So the only organic visitor worth a dollar of effort is one who lives inside those two circles.

The metric is qualified local leads, not sessions. Capacity is roughly 240 to 320 enrollments a year (two locations, three to four cohorts of ~40). That's the ceiling. A page that pulls 5,000 monthly readers from Ohio, Texas, and the UK does nothing for that number, and it actively hurts: it inflates vanity metrics, dilutes the site's local relevance signals, and wastes writing time. Success looks like the local map pack showing Bob top-three across the Front Range, town pages ranking for "[service] [town]," and the phone actually ringing from Boulder County and the Denver metro.

What this means for what we DON'T do: generic national "divorce recovery" informational content is low-value here. The whole national-blog-SEO playbook (rank a big pillar for "how long does divorce recovery take," rack up traffic) optimizes for exactly the reader Bob can't enroll. The three-hub content architecture in 03-content-cluster-architecture.md is real and worth building eventually for brand authority and GEO, but it is NOT the growth engine for filling seats. The leverage is local: the Google local pack, the two Google Business Profiles, town and service pages, local reviews, and local links. National blog volume is a distant second, and only useful to the extent it feeds AI citations and brand prominence that circle back to local trust.

One honest caveat up front: these are genuinely small-volume local queries. That's fine. A "divorce class near me" search in Colorado runs about 10 a month but carries a $10.49 cost-per-click, because the person searching it is ready to enroll. We're not chasing volume. We're chasing the handful of high-intent local people each month who are looking for exactly what Bob runs. (Real numbers in Section 2.)

2. Local keyword and intent map

The real numbers (DataForSEO, Colorado geography, pulled 2026-07-11)

These are Colorado-scoped Google Ads search volumes, not national. They confirm the thesis: low volume, high intent, high commercial value.

How to read this: "divorce recovery Boulder" not returning a measurable number does not mean nobody in Boulder wants this. It means people in Boulder mostly type "divorce support group near me" or "divorce classes near me" and let Google localize. That's why the grid scan (Section 4) matters more than exact-match keyword volume, and why town pages plus a strong GBP beat trying to rank a national keyword.

The intent spectrum, mapped to surfaces

From the seo-future-strategy.md five-query set, sorted by how we capture each:

1. "near me" and un-modified local intent ("divorce support group near me," "divorce class near me," "divorce counseling," "divorce seminar"). Google localizes these to the searcher's location. Captured by the GBP / local pack, not by a keyword-stuffed page. This is the highest-value, highest-leverage bucket.
2. Town-modified service queries ("divorce recovery Boulder," "divorce support group Denver," "divorce counseling Longmont"). Captured by town landing pages that rank in the classic organic (blue-link) results below the map pack.
3. Format and brand-generic ("divorce recovery workshop," "divorce seminar," "Rebuilding workshop," "Fisher divorce adjustment"). Bob already owns most of these organically. Captured by the existing homepage / workshop page, protect and reinforce.
4. Informational, national ("how long does divorce recovery take," "stages of divorce grief," "dating after divorce"). Deprioritize. Build later for GEO and authority only, never as the seat-filling engine. If we build them, they must link down hard to the local pages and the workshop CTA so any rare local reader converts.

Deprioritize explicitly, and why

• National informational pillars: pull out-of-state readers who can't enroll. Low local ROI.
• Broad "divorce" head terms: dominated by legal and national players, wrong intent (people want lawyers, not recovery), and non-local.
• Anything targeting a national audience for its own sake. Every hour spent ranking a national keyword is an hour not spent on a town page or the GBP that actually fills a Niwot or Cherry Creek seat.

3. Site architecture: town and service pages

The principle

Two page types do the local organic work:

• Service pages: one page per thing Bob offers (the divorce recovery workshop, the intro workshop, the graduate community). These carry the format and brand-generic intent and hold the core schema.
• Town / "areas served" pages: one page per town in the catchment, each targeting "[service] [town]" and reinforcing the map pack. A published city page is one of the highest-ROI moves in local SEO: one 2026 case study saw impressions climb 173% and tracked keywords move an average of +34 positions in six weeks after a single city page went live (growthmindedmarketing.com).

Prioritized town backlog

Scored on (within ~60-min drive) x (population) x (affluence / fit for a $1,200 ten-week program). Reconciled with the cities-within-100mi-niwot.md list and the seo-future-strategy.md candidate table. Two catchments, tiered.

Niwot circle (Boulder cohort):

Cherry Creek circle (Denver cohort):

Ship Tier 1 first (Boulder, Longmont, Louisville, Lafayette for Niwot; Denver and Cherry Creek for the second location once that GBP exists). Let the grid scan (Section 4) reorder Tier 2 and 3 by where Bob is actually invisible on the map.

Per-town page structure

Building on the 01-local-seo-checklist.md §6 template, each town page (about 1,000 to 1,400 words, genuinely useful, not thin doorway junk):

• H1: "Divorce recovery workshop serving [Town], CO"
• Opening: drive time and distance to the relevant cohort venue (Niwot or Cherry Creek), and how many graduates have come from that town if any.
• What the workshop actually is: the short version of the Fisher rebuilding framework, one paragraph, linking up to the full workshop page.
• Why it's different: trained-facilitator small-group cohort (Bob, LPC plus a former Google engineer, and Kathy Decker, MA LPC, co-facilitating), not a peer-led drop-in or a church group. Since 1993. 3,000+ alumni.
• Local context that's actually helpful: nearby court self-help resources, local co-parenting resources, linking OUT (those are their keywords, and linking out builds trust).
• Local proof: two or three testimonials from graduates from that town or county when we have them, general 5-star reviews otherwise (Bob has 293 reviews at 4.96 stars to draw from).
• Logistics: price ($1,200, or $1,099 early-bird through Fri July 24, 2026), next cohort start (Sun Sept 20, 2026), the $30 intro workshop as the low-friction first step.
• FAQ block (5 to 7 questions tailored to that town) with FAQPage schema.
• LocalBusiness schema with the correct cohort venue's location for that circle.
• One clear CTA (book the 15-minute call with Bob) with a UTM tag: ?utm_source=town-page&utm_medium=organic&utm_campaign=[town-slug].

Internal linking: every town page links up to a /areas-served/ (or /divorce-recovery-front-range/) hub and across to the workshop page, per the linking rules in 03-content-cluster-architecture.md.

4. Google Business Profile and the local pack

The local pack is the single highest-leverage local-organic surface. In the 2026 Whitespark-style weighting, GBP signals are about 32% of local pack ranking influence, the largest single category, ahead of on-page (19%), reviews (16-20%), links (15%), behavioral (8%), and citations (7%) (w3marketinghub.com, localdominator.co). If we do one thing well, it's this.

Note the 2026 shift: on mobile in the US, Google has started replacing the classic 3-pack with an AI "local pack" that synthesizes a recommendation and surfaces about 32% fewer unique businesses (crawlvision.com). Fewer slots means being in the top three matters more, not less, and it raises the value of review signals and entity clarity that the AI leans on.

Two profiles

• Niwot (primary, exists): already live. Per the decisions log, its primary category is currently set to "Divorce service" (verified live 2026-06-07). Keep, optimize per the full google-business-profile-checklist.md.
• Cherry Creek / Denver (planned second listing): stand up when the Denver cohort location is locked. Its exact street address is deliberately withheld since class locations rotate, so run it as a service-area listing with the address hidden.

The GBP work, both locations

• Service areas: list real towns from the catchment (Google allows up to 20). Do NOT overextend; inflated service areas get detected and de-trusted. Niwot lists the Boulder circle; Cherry Creek lists the Denver circle. Overlap towns (Broomfield, Westminster) can appear on both.
• Address handling: service-area mode, physical address hidden (correct for a rotating-venue business, and Google policy for businesses that don't serve customers at a fixed storefront).
• Posts: 2 per week minimum (the ~30-day decay rule), rotating Update / Offer / Event / Product. Build a 30-post bank up front. Event posts carry the Sept 20 cohort start and match the Event schema. Offer posts carry the $1,099 early-bird and the $30 intro workshop.
• Q&A: pre-seed with the 12 FAQ questions (the kit already exists in the checklist). Answer real questions in Bob's voice within 24 hours. Flag spam monthly.
• Photos: 10+ in launch week, 25+ within 90 days, one new photo a week for the freshness signal. Real cohort-room photos, the 19-blocks pyramid as a labeled image, Bob and Kathy. Age the imagery correctly: the audience is women averaging about 50, so photos should show midlife adults, not people in their 20s.
• Reviews: see Section 5. This is where GBP and rankings compound.

5. Reviews as a ranking asset

Bob's review base is a genuine moat: 293 reviews at 4.96 stars, on top of 3,000+ alumni since 1993. Most local competitors (solo therapists, church groups) have a handful of reviews or none. In 2026 weighting, review signals are 16 to 20% of local pack rank, and review velocity is itself a factor: two fresh reviews a month beats twenty in a burst and then silence (localdominator.co, kaidm.com).

Three ways the review base feeds local rankings:

1. Volume and rating establish prominence, a core local ranking factor. Bob is already far ahead here; the job is to not let it go stale.
2. Review keywords and locations matter. A review that says "the divorce recovery class in Boulder changed things for me" reinforces both the service term and the town. Gently prompt graduates to mention what the workshop was and, naturally, where they came from.
3. Velocity keeps the profile looking alive. A steady drip signals an active, real business to both Google and the AI local pack.

The ongoing loop (simple, consent-first)

• At graduation (every cohort, three to four times a year), Bob or Kathy invites graduates to leave a Google review, using the one-tap review-link shortcut. Ten graduates a cohort leaving reviews is 30 to 40 fresh reviews a year, comfortably above the two-a-month velocity floor, and spread across the year rather than dumped.
• Respond to every review within 48 hours in Bob's voice, the positive ones included (response rate is itself a signal, and it models engagement).
• Ask 5-star reviewers for a photo and permission to show it beside their quote on the site (the checklist §3.8 outreach, run by Sparrow). Faces convert better than anonymous text.
• Cross-list a few reviews to the second-location GBP context once it exists, and to Facebook and Psychology Today, so the review presence isn't single-surface.
• Monitor monthly for fake or competitor-malicious reviews.

Honor the hard rule: graduates' own review text may name Shea if they choose, do not censor it, but nothing we write ever mentions her.

6. Local links and citations

Citations are now only about 7% of direct ranking weight, down from their old prominence, but they still do essential work: they're entity-verification signals that let AI local results and Google confirm Bob is one real, consistent business (kaidm.com, w3marketinghub.com). Quality beats quantity hard: a niche or authoritative citation (BBB, a Front Range chamber, a therapist directory) can carry 10 to 15 times the weight of a generic free directory.

NAP consistency first

Citation drift silently suppresses the local pack. Before chasing new links, get one canonical NAP (name, address-handling, one phone number, ideally a dedicated Google Voice line) identical across GBP, the site footer, the on-site LocalBusiness schema, and every existing directory. The seo-future-strategy.md sequences a Whitespark cleanup (~$80) of the top ~14 directories before any grid scan; keep that order. Then quarterly re-audit the top 30 directories.

Citation and link targets (local, tiered)

• Tier 1, foundational: Google (both profiles), Apple Maps, Bing Places, Yelp, BBB, Facebook. Get NAP identical here first.
• Tier 2, local + niche: Niwot Chamber, Boulder Chamber, Longmont Chamber, a Denver-area chamber for the second location, plus therapist and counseling directories (Psychology Today, GoodTherapy, TherapyDen). These double as GEO corpora (Section 9).
• Local backlinks (the real leverage): these overlap directly with the offline / local-press plan (offline-and-local-press-marketing-plan.md) and the automation blueprint, so they compound instead of being separate work:
  • Local press (Daily Camera, Boulder Weekly, Longmont Times-Call, 5280, Denver Post): a single feature or op-ed on gray divorce earns a high-authority local backlink AND press credibility. The offline plan's Sample 1 pitch and the gray-divorce op-ed are the vehicles.
  • Community calendars (library, city, Nextdoor, Patch): every intro talk and cohort start listed. Low effort, local relevance, some carry links.
  • Front Range referral partners (family-law attorneys, mediators, divorce financial planners, therapists, churches with divorce-care programs): the offline plan's #1 channel. When any of them list Bob as a resource on their own site, that's exactly the local, topically-relevant backlink Google and AI weight most. Ask referral partners who have a website to add a "resources" link.
  • Free intro talks at libraries and community centers: the host often lists the event with a link, and the talk itself is a local-authority signal.
  • Local awards (Best of Boulder, Boulder County Gold): a nomination or win is a durable local trust link.

The rule: don't run link building as a separate SEO project. It IS the offline plan, viewed through an SEO lens. Every warm intro, every press hit, every calendar listing, every talk is also a local ranking signal. Cross-reference the two plans so nobody does the same outreach twice.

7. Content that ranks locally and actually helps

Keep this small and local. The job is not a national content mill.

• Front Range-specific angles that a local person actually searches and a local editor would actually run: "gray divorce on the Front Range," "divorce resources in Boulder County," "what a divorce recovery class in the Denver area looks like." These serve intent and give local press and local links something to point at.
• The graduate community / social-activities angle is a genuine growth lever, and it widens the top of the funnel. People won't drive 90 minutes for weekly class, but they will drive further for a graduate hike, potluck, or social. That means content about the ongoing community (what happens after the ten weeks, the alumni events, the private graduate group Bob already runs) does two things: it draws a wider local audience into Bob's world, and it earns local links and event listings. Treat the social side as both a local content magnet and a link magnet, while class itself stays tightly local.
• The intro workshop ($30) as the content hook. It's the low-friction entry point. Content and calendar listings should push the $30 intro workshop as "try it," which converts local readers into a room without asking for the full $1,200 commitment up front. Never call it free; the $30 exists so people show up.
• Honor the writing rules everywhere: contractions, no em-dashes, plainspoken, concrete, no motivational filler. Bob's voice comes from his newsletters, not blog-speak.

The three-hub national content architecture in 03-content-cluster-architecture.md stays parked as a Phase-3+ authority and GEO play. It's good work, but it does not fill Sept 20. Build the local surfaces first.

8. Technical, schema, and mobile

• Schema, on every relevant page:
  • LocalBusiness (or ProfessionalService) on the homepage, workshop page, and each town page, with the correct venue location per circle and NAP byte-identical to GBP.
  • Event on the workshop and cohort-start content (Sept 20, 2026 start), matching the GBP Event posts.
  • FAQPage on the workshop page and every town page.
  • Person for Bob (LPC, former Google engineer) and Kathy Decker (MA, LPC), the single most important entity page for AI (Section 9).
• Mobile-first is not optional: 43% of current visits are on a phone, on an Apple-heavy audience (GA4 deep-dive). The next site build must be genuinely fast on mobile, tap targets sized, the book-a-call CTA reachable without pinch-zoom.
• Internal linking out of the homepage. Today the homepage takes 73% of all landings and the deeper pages barely get direct entries (GA4 deep-dive). That's a structural gap: the homepage hoards authority and never passes it to the town and service pages that would rank locally. Fix it by linking prominently from the homepage to the areas-served hub and the key town pages, and from every town page back to the workshop CTA.
• Speed and crawlability: clean, fast, no bloat, XML sitemap with lastmod dates (freshness helps AI retrieval), correct canonical tags so town pages don't read as duplicates.

9. AI and GEO, for local

The goal is narrow and local: when someone in Boulder or Denver asks ChatGPT, Gemini, Perplexity, or Google's AI Overview "where can I do a divorce recovery class near me," Bob's name comes back. In 2026 the mobile local pack is increasingly AI-synthesized and surfaces fewer businesses (pinmeto.com), so being the entity the AI trusts is the local game, not a separate national one.

What actually drives local AI recommendation (from 02-llm-seo-strategy.md, validated against current research):

• Entity clarity via schema: Person + Organization + LocalBusiness/ProfessionalService. AI engines strongly prefer well-typed sources. The About-Bob page with full Person schema is the highest-value AI page on the site.
• Consistent NAP across the open web: this is now literally entity-verification for AI. The citation cleanup in Section 6 doubles as GEO work.
• Review signals: the 293/4.96 base and steady velocity are exactly the kind of prominence signal AI local answers lean on.
• Presence in the corpora AI cites: Psychology Today, GoodTherapy, therapist directories, plus authentic local Reddit and Quora participation ("anyone done a divorce recovery group in the Denver area?"). These are high-trust sources AI triangulates from.
• FAQ-format content phrased the way locals ask, extractable straight into AI answers.

Do this as one workstream with the schema and citation work; the per-engine differences are small relative to the cost of doing it once.

10. Measurement: local leads, not sessions

The metric is qualified local leads and cohort fill. Two things block clean measurement today, and one is urgent.

KPIs to watch, once tracking exists:

• Local map-pack rank: the Local Falcon grid scan (Section 12) across the five generic queries, measured as % of grid points where Bob is top-three, tracked over time and after each new town page or citation cleanup.
• Town-page organic rank: "[service] [town]" positions in SE Ranking (or Search Console) for the Tier 1 towns.
• GBP actions: calls, direction requests, website clicks, and booking-link clicks from each profile. These are the closest proxy to a local lead we have even before full conversion tracking, so start reading them now.
• Bookings and enrollments by source: the actual business metric, via the UTM tags on town-page and GBP CTAs plus the booking system, once conversion events fire.
• Cohort fill rate and lead-to-seat: are the Niwot and Cherry Creek cohorts filling their ~40 seats, and how many leads did it take.

Explicitly not a KPI: total sessions, national pageviews, or "traffic up X%." A traffic number that includes out-of-state readers is noise here. Filter known bots and internal traffic on the next setup and discount current totals 10 to 20% for bot inflation.

11. Roadmap

Sequence so the highest-leverage local surfaces (the local pack, GBP, first town pages) come first, and nothing waits on the big content build.

90-day sprint (highest leverage first)

12-month arc

• Months 4-6: stand up the Cherry Creek / Denver second GBP once the location is set; write the Denver-circle Tier 1 town pages (Denver, Cherry Creek); run a Denver-centered grid scan. Continue review velocity and 2x/week posts on both profiles.
• Months 6-9: Tier 2 and 3 town pages across both circles, reordered by grid data. Deepen the therapist/attorney referral link base. Begin light Front Range-specific and graduate-community content.
• Months 9-12: GEO push (directory corpora, forum presence, FAQ content), local-awards cycle, and only now, if capacity allows, begin the national content hubs from 03-content-cluster-architecture.md for authority and AI citations, always linking down to local.

12. Tooling

Frame the paid stack as optional and only when we're ready to act on what it tells us. Nothing here blocks the free foundation work.

Free, do now

• Google Search Console (both properties eventually), Google Business Profile Insights, GA4 (once conversion events exist).
• Screaming Frog (free tier) for the site crawl. GMBSpy Chrome extension to spot-check competitors' GBP categories.

Paid, when ready (~$81/mo recurring, per seo-future-strategy.md)

• SE Ranking Essential (~$52/mo): keyword rank tracking for "[service] [town]" and backlink/competitor gap. Start on the free trial.
• Local Falcon (~$29/mo): the grid scan, the single most useful local tool here. Baseline plus scheduled re-scans. Roughly $4.50 per full five-keyword sweep.
• One-time: Whitespark NAP cleanup (~$80).

Sequencing (locked in the source file): SE Ranking trial first, Whitespark cleanup before any grid scan, Local Falcon after cleanup, re-scan at day 45 for lift.

13. Platform portability (Rebuilding may leave GoHighLevel)

Annette has flagged that Rebuilding may move off GoHighLevel soon. So this plan is deliberately platform-agnostic, and we should NOT sink effort into GHL-specific SEO plumbing that a migration would throw away. In particular, do not build conversion tracking into GoHighLevel if a move is likely; scope it for the next platform instead.

Treat the site rebuild as the chance to get the SEO architecture right from the start. If and when the site moves, these are launch-day non-negotiables, not later cleanup:

• Clean, stable URLs: /areas-served/[town]/, /workshop/, /about/, no CMS-generated junk, and 301 redirects mapping every old URL to its new home so nothing loses its existing rank (Bob already ranks #1 for "divorce recovery workshop," protect it).
• Town-page architecture built in: the areas-served hub and town-page template as first-class page types, not bolted-on blog posts.
• Full schema support: the platform must let us inject LocalBusiness, Event, FAQPage, and Person JSON-LD per page.
• Fast, mobile-first rendering: 43% of visits are mobile; the new site has to be quick on a phone.
• GA4 conversion tracking wired from day one: form submit, phone tap, booking, intro-workshop signup, all as key events. This ends the two-year blind spot and is the biggest single measurement unlock.
• NAP and canonical consistency: one NAP, correct canonicals, sitemap with lastmod.

Get these six right at launch and the local plan runs on top of a foundation that won't need re-doing.

Open decisions for Annette / Bob