How a Roofing Contractor Closed the AI Search Gap and Started Showing Up in Gemini
A roofing contractor in a competitive local market was running Google Ads, had an active Google Business Profile, and ranked reasonably well in traditional sear...
Cody Vincent
Chief Revenue Officer
A roofing contractor in a competitive local market was running Google Ads, had an active Google Business Profile, and ranked reasonably well in traditional search. Leads were slowing anyway. A competitor was showing up in Gemini and ChatGPT answers. They were not.
That gap is not a mystery. It is a set of specific, fixable problems. This article walks through what those problems were, how they got fixed, and what changed after.
If you run a contracting business and you are wondering why AI search keeps skipping your brand, this is the article for you.
The Problem Was Not the Website. It Was What the Website Was Missing.
Most contractors have a website. Most of those websites share the same structural problems from an AI-visibility standpoint.
This contractor had a site with decent photos, a phone number, and a few service pages. By traditional SEO standards, acceptable. By AI-readiness standards, nearly invisible.
The audit surfaced these specific gaps:
- Thin service pages with no depth, no FAQs, and no structured information AI systems could extract and cite
- Missing schema markup — no LocalBusiness schema, no Service schema, no Review schema
- No llms.txt file — a simple but important signal that tells AI crawlers what content is available and how to interpret it
- Weak trust signals — inconsistent NAP (name, address, phone) data across directories, sparse review volume, and no third-party citations pointing back to the site
- Listing problems — incomplete or conflicting information on Google Business Profile and key directories
None of these are exotic problems. Every one of them is a reason AI systems like Gemini, ChatGPT, and Perplexity pass over a business when generating an answer.
Why AI Systems Skip Contractors
When someone types "best roofing contractor near me" into Gemini or asks ChatGPT for a recommendation, those systems are not scrolling through websites the way a human does. They pull from structured, crawlable, verifiable information.
Thin service pages give them nothing to cite. Missing schema means the AI cannot confirm what you do, where you operate, or how to categorize you. Weak trust signals mean you lose to the competitor with 80 reviews and consistent directory listings.
Your competitor is not better at roofing. They are just easier for AI to cite.
That is the core of AI search engine optimization — making your business legible to both human buyers and AI systems at the same time. The technical requirements overlap more than most contractors realize.
What the Audit Found: A Score of 34 Out of 100
The contractor ran a free scan through New Reward. The result was a readiness score of 34 out of 100.
That number is not a grade. It is a ranked list of gaps. The audit broke the score down into specific, named problems — not vague recommendations like "improve your content," but actionable items like "your roofing service page has under 300 words and no FAQ schema" and "your llms.txt file does not exist."
The score covered every surface that matters: Google, Google AI Overviews, ChatGPT, Perplexity, Gemini, Claude, and Grok. Not just one engine. All of them.
That matters because AI visibility is not a single channel. A buyer might ask ChatGPT for a recommendation in the morning and check Gemini in the afternoon. Optimizing for one while ignoring the others still leaves you invisible.
The Fixes That Were Shipped
This is where the process differs from what most tools offer. Most AI-visibility tools stop at the audit. They hand back a report and expect the client to execute. For a roofing contractor managing a crew and juggling jobs, that report sits in an inbox.
New Reward shipped the fixes directly. Here is what changed.
Service Page Depth
Each core service page — roof replacement, roof repair, emergency tarping, gutter installation — was rebuilt with structured content. Clear service descriptions, local context (city and county references), FAQ sections with questions buyers actually ask, and schema markup embedded at the page level.
The goal was not word count. The goal was making each page something an AI system could read, understand, and cite.
Schema Markup
LocalBusiness schema, Service schema, and Review schema were added across the site. This tells AI crawlers exactly what the business does, where it operates, what its hours are, and how buyers have rated it. Without schema, a site is readable to humans but largely opaque to AI systems.
llms.txt
A properly structured llms.txt file was created and deployed. This file signals to AI systems what content on the site is available for training and citation. Most contractor websites have never heard of it. That absence is a gap competitors can exploit.
Trust Signal Cleanup
NAP data was audited and corrected across Google Business Profile, Yelp, Angi, HomeAdvisor, and a set of regional directories. Inconsistent listings confuse AI systems trying to verify a business's legitimacy and location. Consistency is a basic trust signal most contractors overlook.
Review Velocity
The contractor had 22 Google reviews at the time of the audit. The Review Velocity Engine — a structured process for generating a steady flow of new reviews — was activated. Within the engagement period, that number moved meaningfully. Review volume and recency are direct inputs into how AI systems assess business credibility.
What Changed After the Fixes
Before-and-after evidence is not optional. Every change was documented so the contractor could see exactly what moved and why.
The practical outcome: the business began appearing in Gemini answers for local roofing queries. ChatGPT started citing the business when asked for contractor recommendations in the service area. Google AI Overviews picked up FAQ content from the rebuilt service pages.
The readiness score moved from 34 to 71 over the course of the engagement.
That improvement reflects specific, documented changes — not a general sense that things got better. If you want to understand what GEO (generative engine optimization) measurement actually looks like in practice, GEO now has a measurement layer that makes before-and-after accountability possible in a way it was not 18 months ago.
Why This Matters for Contractors Specifically
Contracting is a high-intent, high-ticket local category. When someone asks Gemini for a roofing contractor, they are not browsing. They have a problem and they need someone now.
That buying behavior is shifting toward AI-generated answers faster than most contractors realize. The businesses showing up in those answers are not necessarily the best contractors in the market. They are the ones whose public footprint is structured well enough for AI systems to read and cite.
SEO, AEO (answer engine optimization), and GEO for local businesses are not three separate strategies. They are three requirements for the same outcome: being findable when a buyer is ready to hire.
Most contractors are investing in one of the three and ignoring the other two. That is the gap.
The Execution Problem Most Tools Do Not Solve
Here is an honest look at the current tool landscape for AI visibility.
Tools like Otterly.ai, Semrush's AI Toolkit, and Ahrefs Brand Radar are monitoring products. They track whether your brand appears in AI answers and surface gaps. They do not fix anything. The execution burden stays with you or your team.
Profound, the category leader, starts at $399 to $499 per month for multi-engine coverage and does not ship fixes. Ahrefs Brand Radar adds $699 per month on top of an existing Ahrefs subscription, with no execution capability. Scrunch starts at roughly $300 per month and is advisory only.
For a roofing contractor, none of those tools solves the actual problem. Knowing you have a gap is not the same as closing it.
New Reward scores, fixes, and proves. The free scan takes roughly 60 seconds and requires only an email address. The audit that follows is specific and ranked. The fixes are shipped by the New Reward team — not handed back as a to-do list. Every change comes with inspectable before-and-after evidence.
Scan, score, fix, evidence. That is the loop.
What AI Visibility Maintenance Looks Like After the Initial Fix
Closing the gap is not a one-time event. AI systems update their training data, new competitors enter the market, and your service pages need to stay current.
What SEO maintenance looks like in 2026 has changed significantly. It is not just crawl monitoring and link audits anymore. It includes keeping schema current, maintaining review velocity, updating llms.txt as your services change, and tracking whether your citations in AI answers hold or slip.
For contractors, the practical implication is straightforward: AI visibility is an operating layer, not a project. The businesses that stay visible are the ones treating it that way.
Get Your Free AI-Visibility Score
If you run a contracting business and you are not showing up in Gemini, ChatGPT, or Perplexity answers, the first step is knowing your score.
Run a free scan at Newreward.com. It takes roughly 60 seconds, requires only your email address, and produces a 0–100 readiness score with a ranked list of specific gaps.
No credit card. No report that sits in your inbox. If you decide to move forward, the team ships the fixes and documents what changed.
Your competitor is showing up in AI answers. You should be too.
Frequently Asked Questions
What is AI search visibility for contractors? AI search visibility means your contracting business appears when buyers ask AI systems like Gemini, ChatGPT, or Perplexity for local service recommendations. It requires structured service pages, schema markup, consistent directory listings, and trust signals that AI systems can read and cite — not just a website that looks good to human visitors.
Why is my roofing company not showing up in Gemini or ChatGPT? The most common reasons are thin service pages with no structured content, missing schema markup, no llms.txt file, inconsistent NAP data across directories, and low review volume or recency. These are specific, fixable gaps — not general website quality issues.
What is schema markup and why does it matter for contractors? Schema markup is structured data added to your website that tells AI systems and search engines exactly what your business does, where it operates, and how buyers have rated it. Without it, AI systems cannot reliably categorize or cite your business in generated answers. LocalBusiness, Service, and Review schema are the three most important types for contractors.
What is an llms.txt file? An llms.txt file sits at the root of your website and signals to AI crawlers what content is available and how to interpret it. It is relatively new but increasingly important for AI visibility. Most contractor websites do not have one — and that absence is a gap competitors can exploit.
How long does it take to start appearing in AI-generated answers after fixes are made? There is no guaranteed timeline, and any specific number would be speculative. AI systems update their data at different cadences. Structural fixes like schema and llms.txt tend to be picked up faster than content authority signals like review volume. The roofing contractor in this case study saw measurable score improvement and AI-answer appearances within the engagement period.
What is the difference between monitoring AI visibility and fixing it? Monitoring tools like Otterly.ai, Semrush's AI Toolkit, and Ahrefs Brand Radar track whether your brand appears in AI answers and surface gaps. They do not fix anything. New Reward scores your visibility, ships approved fixes on your behalf, and documents before-and-after evidence of what changed. Knowing the gap exists is not the same as closing it.
How do I get started with AI visibility for my contracting business? Run a free scan at Newreward.com. It takes roughly 60 seconds, requires only your email address, and produces a 0–100 readiness score with a ranked audit of specific gaps. If you decide to move forward, the team handles execution and provides inspectable evidence of every change.