Note: This is an illustrative case study based on real GEO optimization techniques and realistic outcomes. "South Bay Plumbing" is a composite example created to demonstrate what a typical local business GEO journey looks like. The specific steps, timeline, and score improvements reflect what we have observed across our work, not a single verified client engagement.
Meet South Bay Plumbing. Family-owned, 15 years in Hermosa Beach. Great reviews on Google (4.7 stars, 180+ reviews). A solid website. Plenty of work from word-of-mouth referrals and Google search. By every traditional metric, a successful small business.
But when we ran their AI visibility scan, the result was stark.
Not a single AI model mentioned South Bay Plumbing when asked "best plumber in Hermosa Beach." ChatGPT recommended two competitors. Claude said it could not recommend specific businesses. Perplexity listed three plumbers from Yelp, none of them South Bay. Gemini pulled from Google Maps but showed plumbers from Torrance and Redondo Beach first.
A business with 15 years of reputation, 180 Google reviews, and a 4.7-star rating was completely invisible to AI search. Here is how that changed over 6 weeks.
The Starting Point Audit
Before fixing anything, we needed to understand why South Bay Plumbing was invisible. The audit found several gaps:
| Platform | Status at Start |
|---|---|
| Google Business Profile | Claimed but incomplete. No services listed, no FAQ, minimal description, photos from 2021. |
| Foursquare | Not claimed. Auto-generated listing with wrong phone number and no business hours. |
| Bing Places | Not claimed. No listing existed. |
| Website Schema markup | None. Plain HTML with no structured data. |
| Directory presence | Yelp (claimed), Angi (auto-listed), HomeAdvisor (expired). That was it. |
| NAP consistency | Phone number wrong on 3 platforms. Address format inconsistent across all. |
The diagnosis was clear: South Bay Plumbing had built a strong Google presence but had done nothing for the platforms that feed AI models. No Foursquare, no Bing Places, no Schema markup, and inconsistent data across directories.
Week 1-2: Foundation Work
WEEKS 1-2
Claiming platforms and fixing data
The first two weeks focused entirely on building the foundation that AI models read:
- Foursquare: Claimed the listing, corrected the phone number, added business hours, uploaded 8 photos, wrote a complete business description, and added all service categories.
- Bing Places: Created a new listing from scratch. Added services, hours, photos, and a detailed business description.
- Google Business Profile: Rewrote the description to include specific services. Added 12 new photos. Listed all services with descriptions. Created 5 FAQ entries answering common customer questions. Added the service area.
- NAP cleanup: Standardized the business name, address, and phone number format across every platform. Fixed the 3 listings with wrong phone numbers.
Week 2-3: Schema Markup and Content
WEEKS 2-3
Making the website AI-readable
- Schema.org markup: Added LocalBusiness schema to the homepage with complete business info (name, address, phone, hours, geo coordinates, service area, price range). Added Service schema for each service page. Added Review schema for testimonials.
- FAQ content: Created a detailed FAQ page answering 15 real customer questions: "How much does a water heater replacement cost in Hermosa Beach?", "Do you offer emergency plumbing on weekends?", "What areas do you serve?" Each answer was detailed and specific.
- llms.txt file: Created and deployed to the website root with complete business information.
- Google Search Console: Submitted the new FAQ page and Schema-enhanced pages for indexing. Requested re-crawl of the homepage.
Week 3-4: Directory Expansion and Reviews
WEEKS 3-4
Expanding presence and social proof
- Directory submissions: Submitted to 12 additional directories beyond the big three. Industry-specific directories (HomeAdvisor, Thumbtack, Porch), general directories (YellowPages, BBB, CitySearch), and Foursquare (which AI models actually read).
- Review campaign: Started asking recent customers to leave reviews on Google and Yelp. Not fake reviews. Not incentivized reviews. Simple follow-up texts after completed jobs asking happy customers to share their experience. Over the next 3 weeks, this generated 15 new reviews across Google (10), Yelp (3), and Foursquare (2).
- Citation building: Made sure every directory listing had the same business name format, same address format, and same phone number. Consistency across citations is one of the strongest signals for both Google and AI models.
Week 4-5: First Results
WEEK 4
AI models start noticing
Four weeks in, we ran the scan again.
The changes:
- Gemini started recommending South Bay Plumbing for "plumber in Hermosa Beach" queries, pulling from the updated Google Business Profile.
- ChatGPT mentioned them in some queries but not consistently. The Foursquare data had been indexed, but ChatGPT was still favoring competitors with longer-established Foursquare profiles.
- Claude was still not mentioning them by name.
- Perplexity started citing them when pulling from Yelp results.
Partial progress. The foundation was working, but the AI models needed more signals to consistently recommend them over established competitors.
Week 5-6: Pushing to Recommended Status
WEEKS 5-6
Content depth and review momentum
- Additional FAQ content: Added 10 more FAQ entries focusing on specific queries AI users would ask: "emergency plumber Hermosa Beach," "water heater repair South Bay," "plumber for old homes Manhattan Beach."
- Blog content: Published 2 articles on the website: "5 Signs You Need a Water Heater Replacement" and "Emergency Plumbing: What Counts as an Emergency?" Both included Schema markup and were submitted to Google Search Console.
- Review momentum continued: 15 new reviews now live across platforms. The Foursquare listing had grown from 0 to 2 reviews and a claimed profile with complete information.
- Foursquare tips and photos: Added additional tips and updated photos. Engaged with the 2 Foursquare reviews.
Week 6: Final Scan
Final AI visibility score after 6 weeks. South Bay Plumbing is now recommended by ChatGPT, Gemini, and Perplexity for plumbing queries in the South Bay area. Claude mentions them when asked directly but does not yet include them in category recommendations.
What Made the Biggest Difference
Looking back at the 6-week timeline, the actions that had the most impact were, in order:
- Claiming Foursquare - This single action is what made ChatGPT start mentioning them. Before claiming, they did not exist in ChatGPT's primary local data source.
- GBP optimization - The detailed services, FAQ, and fresh photos made Gemini recommend them almost immediately.
- Schema markup - This helped across all AI models by making the website's data machine-readable.
- Bing Places - Important for ChatGPT's web search component. A claimed Bing listing appeared in ChatGPT's search results.
- New reviews - The 15 new reviews provided fresh social proof. AI models weight recent reviews more than old ones.
The directory submissions and llms.txt file were good hygiene but did not appear to drive measurable score changes on their own.
The Cost Breakdown
| Action | Cost | Time |
|---|---|---|
| Foursquare claim + optimize | $0 | 45 minutes |
| Bing Places claim | $0 | 30 minutes |
| GBP optimization | $0 | 2 hours |
| Schema markup | $0 (DIY) or $200-500 (developer) | 3-4 hours DIY |
| FAQ content | $0 (DIY) or $150-300 (writer) | 4-5 hours DIY |
| Directory submissions | $0 (most are free) | 3-4 hours |
| Review campaign | $0 | 30 minutes setup + ongoing |
| Total (DIY) | $0 | ~15 hours over 6 weeks |
| Total (PACO GEO) | $149 one-time | ~2 hours (your time for reviews + approvals) |
What This Means for Your Business
South Bay Plumbing is illustrative, but the pattern is real. Businesses with strong Google presence but no Foursquare, no Bing Places, and no Schema markup are invisible to AI search. The fix is specific, finite, and affordable.
The question is not whether to do this work. It is whether to do it now while your competitors are still ignoring AI search, or later when everyone has caught on and the advantage has shrunk.
Want to see where you stand? Run a free AI visibility scan to get your own score. It takes 60 seconds and shows exactly what ChatGPT, Claude, Perplexity, and Gemini say about your business. If the result looks like South Bay Plumbing's starting point, the GEO checklist shows you exactly what to do.
Key Takeaways
- A strong Google presence does not automatically mean AI visibility. AI models use different data sources.
- The three highest-impact platforms are Foursquare, Google Business Profile, and Bing Places.
- Results can appear within 2-4 weeks. Significant improvement in 6 weeks is realistic.
- The total cost can be $0 if you do it yourself, or $149 one-time with PACO GEO.
- Reviews matter. Not hundreds. 15 new reviews across platforms made a measurable difference.
- Schema markup is the technical piece most businesses skip. It is also one of the most impactful.
Get your AI visibility score
Free scan. No signup. See what AI models say about your business right now and find out where you stand.
Check My Visibility Free