A 22-Year Plumbing Company Was Invisible in AI Answers — Even Though Locals Already Trusted Them

Composite case study. Built from the patterns we see across plumbing engagements. Numbers and details are abstracted to protect operator privacy. Real, named work appears here as engagements mature and operators sign off on what we share.

Starting state

The operator was not new. That was the frustrating part.

They had been in business for more than two decades, had strong repeat customers, solid word-of-mouth, marked trucks, and a real local reputation. In the field, they were known. Online, they looked smaller than they were.

When we ran the initial search and AI visibility check, the gap was obvious. They were showing up inconsistently across Google and Maps for broad plumbing terms, but almost never appeared when AI tools were asked buyer-intent questions like:

• "Who should I call for emergency plumbing in [city]?"
• "Best plumber near me for water heater replacement"
• "Reliable drain cleaning company in [city]"
• "Which plumbing company has the best reviews in [city]?"
• "Who handles slab leak detection near [city]?"

The company had real authority, but the internet did not understand it cleanly.

Baseline findings

Signal	Starting Point
AI mention share across core service prompts	4%
Prompts where competitor was named first	68%
Google Business Profile primary category	Correct
Service-area clarity	Weak
Dedicated service pages	6
Service pages with strong local proof	1
Review count	180+
Review velocity	Inconsistent
Calls tracked by source	No
Map Pack average position across priority terms	7.8
Schema coverage	Basic organization schema only
Citation consistency	Mixed NAP variations across directories

The biggest surprise was not that they lacked reviews. They had reviews. The problem was that the proof was scattered. Their reviews mentioned "honest," "fast," and "friendly," but their website did not clearly connect those trust signals to the services AI and buyers were asking about. Their Google profile listed services, but the site had thin pages for several profitable categories. Their city targeting was implied, not structured. Their citations had old phone-number and address inconsistencies from previous vendors.

The business was real. The footprint was messy.

The diagnosis

The operator came in thinking this was a ranking problem. It was not. It was an alignment problem.

Their website, Google profile, citations, reviews, service pages, and local mentions were all telling slightly different versions of the business. That matters more now because AI tools do not "believe" a business just because the homepage says it is trusted. They look for repeated, verifiable signals.

That became the engagement focus.

What Axis37 changed

We did not start by publishing twenty blog posts. We started by cleaning the foundation.

1. Rebuilt the service hierarchy

The old site had a generic "Services" page with a few subpages. It was not enough. We rebuilt the structure around the jobs the operator actually wanted more of: emergency plumbing, drain cleaning, water heater repair, water heater replacement, leak detection, slab leak repair, sewer line repair, garbage disposal repair, fixture installation.

Each service page had a defined job: explain the problem in plain homeowner language, show when to call, explain the operator's process, add city/service-area relevance without stuffing, include review language where available, and tie the page back to conversion — call, request service, or book.

2. Tightened local market signals

The operator served multiple cities, but the site treated the whole region as one blob. We created clearer service-area architecture and internal links so Google, Maps, and AI systems could associate the company with specific cities and specific jobs.

This included priority city references on core service pages, internal links from service pages to service-area pages, local proof blocks, review excerpts tied to city/service combinations, and clearer footer and contact-area structure.

3. Cleaned up citations and entity consistency

Old directory listings had inconsistent phone numbers, old descriptions, missing service categories, or generic business summaries. We cleaned the most visible citation layer first: name/address/phone consistency, category consistency, description alignment, service list cleanup, profile completion, link targets.

The goal was not "more citations." The goal was a cleaner entity footprint.

4. Added schema that matched the actual business

The site had basic schema, but it was not doing enough work. We added or improved: LocalBusiness / PlumbingBusiness schema, Service schema, FAQ schema where appropriate, Review/reference markup where compliant, area served fields, sameAs links, and internal entity consistency.

This was not treated as a magic trick. It was treated as structure.

5. Built a monthly Recommendation Report cycle

Each month we tested the same categories of buyer-intent questions across AI/search surfaces. The report answered: was the company named? Which competitors were named? Why did AI appear to trust those competitors? Which signals were missing? What changed from the prior month? What should we fix next?

This turned AI search from a vague trend into an operating cycle.

6. Added call tracking without breaking local trust

The operator had no clear way to know which service pages or channels were creating calls. We implemented source-level tracking carefully, preserving local trust and avoiding messy NAP issues. Calls were then grouped by Google Business Profile, organic service pages, paid traffic (where applicable), direct, referral, and AI-assisted / dark attribution indicators where trackable.

90-day results

Signal	Start	Day 90
Footprint Score	31 / 100	67 / 100
AI mention share across core service prompts	4%	27%
Prompts where company appeared in top 3 recommendations	8%	34%
Average Map Pack position across priority terms	7.8	4.1
Dedicated service pages	6	14
Priority pages with conversion tracking	0	14
Calls attributed to organic/search visibility	Not tracked	46/month
Calls from priority service pages	Not tracked	19/month
Citation consistency issues	31 found	7 remaining
GBP actions from search/maps	Baseline established	+22% vs. baseline month

The most important change was not a single keyword jump. The important change was that the operator started appearing in the recommendation set.

Before the engagement, AI tools often named larger competitors, lead-gen directories, or companies with cleaner online footprints. After the foundation and authority work, the company began appearing in answers for emergency plumbing, water heater replacement, and drain cleaning prompts.

What surprised the operator

The owner expected the biggest issue to be review count. It was not. The operator already had enough reviews to be credible. The problem was that the reviews were not connected to a structured service footprint.

They had proof. They did not have organized proof.

Single takeaway

For local plumbing companies, AI visibility is not won by publishing generic plumbing articles. It is won by making the business easier to verify.

The company with the clearest service footprint, strongest local proof, cleanest profile data, and most consistent trust signals is the company AI is more likely to name when the buyer asks who to call.

The business was better than its footprint. Our job was to close that gap.

If you want to know what AI actually says about your operation today, run the Search Checkup — it surfaces the same signals we would audit on day one.