You spent $3,000 on Google Ads last month and got 47 calls. Your organic search traffic? Fourteen visits. Fourteen. Meanwhile, your Google Business Profile sits there with a 4.8 star rating, detailed service descriptions, and photos of your team—and still nobody finds you.
Here’s what’s happening: You’re invisible to Google’s new AI search layer, and frankly, most SEO advice from 2022 is making it worse.
A few months back, I watched Google’s Search Generative Experience (SGE) pull an answer about emergency plumbing from a competitor’s site—not because they ranked #1, but because their content directly answered the exact question someone typed. The user never clicked their link. They got the answer, saw a phone number in the snapshot, and called them directly. My client, who had better rankings than the competitor, got nothing.
That’s the new game. And if you’re still chasing “plumber near me” keywords like it’s 2019, you’re fighting yesterday’s battle.
The Death of the Keyword: Why Google’s AI Doesn’t Care About ‘Plumber Near Me’
Let me be direct: keyword research isn’t dead, but the way you’ve been doing it is.
For years, SEO meant matching search terms. You’d rank for “emergency plumber Denver” or “roof leak repair Fort Collins,” and traffic would flow. Google had a simple job—match words in your content to words people searched for. Done.
SGE changed the rules entirely.
When someone searches “why is my AC making a hissing noise” or “how much does a foundation inspection cost,” Google’s AI doesn’t just scan your website for matching keywords anymore. It reads your content, understands what problem you’re solving, compares it against hundreds of other answers, and then generates a summary without your site even appearing as a clickable link.
The user gets their answer. Google gets engagement. You get… ignored.
Here’s the shift that matters: Google moved from keyword matching to intent understanding. And more importantly, they moved to summarization. The AI is looking at your entire body of content and deciding: “Does this business actually understand what they’re talking about?” If yes, your information gets pulled into the snapshot. If no, you’re invisible.
I watched this play out with a local HVAC company in Austin. Their old site had optimized pages for keywords like “HVAC maintenance services Austin” and “air conditioning repair near me.” Rankings? Solid—top 5 for most terms. But SGE summaries? They never appeared. Why? Because the content was thin. It was keyword-stuffed descriptions without real substance.
We rebuilt their approach around answering specific problems: “What does an HVAC maintenance check actually include?” (with photos of the tech checking refrigerant levels, cleaning coils, testing thermostats). “Why your AC bills skyrocketed in July” (with actual cost breakdowns). “When to replace vs. repair” (with a decision tree).
Within 6 weeks, they started appearing in AI summaries. The calls increased 34%. Rankings stayed roughly the same, but visibility shifted to places that actually converted.
The reason? SGE pulls from content that directly answers questions. Not content that vaguely mentions them.
The ‘Helpful Content’ Checklist: What the AI is Looking For
Google published guidelines about “helpful content” for a reason. It’s not vague corporate speak—it’s literally the criteria their AI uses to decide whose answer to surface.
Here’s what actually works:
Transparency in pricing and service areas
Stop hiding your prices. Seriously. I know you think it makes you sound exclusive or gives your sales team leverage, but it does the opposite in the AI era.
When you write “Call for pricing” or “Custom quote needed,” Google’s AI can’t pull that into a summary. It looks like you’re dodging the question. But when you write “Our standard AC tune-up is $149 and includes a 23-point inspection, refrigerant level check, and filter replacement” (or whatever you actually do), the AI has something real to work with. It can surface that directly. The user sees a clear answer. You get the lead.
Same with service areas. If you serve “Greater Denver Metro Area,” you’ve told the AI nothing. If you write “We service Denver, Aurora, Littleton, Boulder, and surrounding areas within 20 miles of our shop,” now the AI can match that against local search intent. It’s specific. It’s verifiable.
For a roofing company in Georgia, this meant completely reframing their pricing. Instead of vague ranges (“roof replacement $5,000-$15,000”), they started with: “Architectural shingle roof replacement on a 2,000 sq ft home typically runs $8,500-$11,000, depending on complexity and location in the Atlanta area. Steep slopes and valleys add roughly $1,000 per section.”
Detailed. Real. Surfaceable.
Specific troubleshooting advice that proves your expertise
This is where most local service businesses fall short. You know this stuff. But your website reads like a generic template.
Instead of “We specialize in HVAC repair,” write this: “If your furnace is running but not heating, the issue is usually a clogged air filter, a broken thermostat wire, or a failed ignition sequence. Here’s how to troubleshoot each one.”
Then actually walk through it. Show photos. Name specific parts. Use real scenarios.
A dentist I worked with stopped writing “We offer cosmetic dentistry” and started writing “Yellow teeth can come from staining (coffee, wine, smoking), internal discoloration (aging, medications, trauma), or enamel erosion. Whitening works best on surface stains but is useless on discoloration caused by tetracycline. Here’s when bleaching vs. veneers make sense.”
Suddenly, patients found them when searching for specific problems. The AI quoted their explanations. They became the authority in the search results.
The impact of ‘Direct Answer’ sections on AI surfacing
Here’s a tactical thing nobody talks about: the structure of your content matters more than the volume.
When you write a blog post with an H2 titled “Does my roof need replacing?” followed by 2-3 clear paragraphs that directly answer it, Google’s AI can extract that section and use it in a summary. The user sees your answer highlighted. You get credit.
But if you bury the answer in paragraph 7 of a 2,000-word post, or worse, if you never directly state an answer and just dance around the topic, the AI can’t easily surface it. It’s ambiguous.
This is why FAQ pages actually do matter now—but only if they’re real and specific, not generic corporate fluff. “What services do you offer?” is useless. “Does your plumbing service handle frozen pipes in the winter?” with a direct answer and a phone number is everything.
Structuring Your Website for the ‘Answer Engine’ Era
Here’s where the tactical implementation kicks in.
Replacing generic FAQ pages with high-value ‘Problem/Solution’ hubs
Stop writing FAQ pages. Start building problem libraries.
Your current FAQ probably looks like: “What is your service area?” “Do you offer emergency service?” “Do you take credit cards?”
Instead, map out the actual problems your customers face. For a plumber, that’s not “what’s your service area”—that’s “Why is my toilet running constantly?” or “What causes a burst pipe?” or “How much does it cost to unclog a main line?”
Then create a dedicated section (call it a “Common Issues” hub or “Troubleshooting Guide”) where each problem gets its own page or detailed section with:
- A clear explanation of what causes the problem
- DIY troubleshooting steps (be honest if they won’t work)
- What professional service involves
- Rough cost range
- A “Schedule Service” link
A roofer in Colorado built this for hail damage. Instead of a generic FAQ, they created pages titled “Hail Damage Signs You Might Miss,” “How to File an Insurance Claim for Hail Damage,” “Hail Damage vs. Normal Wear and Tear,” “DIY Roof Inspection After a Storm.”
Each page is 800-1200 words, deeply helpful, and directly answering what someone’s actually searching for after a hail storm. They appear in SGE summaries constantly now. Their phone rings.
Using micro-data (Schema) to tell Google exactly what questions you answer
This is the technical piece that separates the people who actually understand AI search from people just guessing.
Schema markup is code that tells Google what your content is about, before Google even reads it. There are specific Schema types for services, FAQs, HowTo content, and local business information.
If you implement FAQ Schema on your “Common Issues” hub, you’re literally telling Google: “This is an FAQ page with real questions and answers.” Google then extracts those Q&As and uses them in summaries and featured snippets.
If you use LocalBusiness Schema with detailed pricing, service area, and availability information, you’re feeding the AI structured data it can trust and surface.
Most local service websites don’t use Schema at all, or use it poorly. If you’re at least implementing basic LocalBusiness Schema plus FAQ Schema on your troubleshooting pages, you’re already ahead of 90% of your competition.
A plumbing company in Phoenix added Schema markup to their emergency service page. Within two weeks, their answer was showing up in local SGE summaries alongside their booking link. Simple markup. Massive impact.
Why ‘short and punchy’ wins over long, rambling blog posts in AI summaries
Here’s the thing: long-form content is great for ranking in traditional search. But for SGE summaries, conciseness wins.
Google’s AI can’t easily extract a direct answer from a 3,000-word blog post. It can extract a single, clear 200-word section that directly answers a specific question.
This doesn’t mean your pages need to be short. It means they need to be structured. Each section should answer one specific question clearly. Use headings that mirror actual search queries. Keep paragraphs short and direct.
A dentist’s page on “does teeth whitening hurt” shouldn’t be 2,000 words of rambling. It should be:
Does teeth whitening hurt?
Whitening can cause temporary sensitivity, especially if your enamel is already thin. Here’s what happens: The bleaching agent opens pores in your enamel to reach the dentin layer beneath, which contains more nerve endings. You’ll typically feel sensitivity to cold for 24-48 hours after treatment.
We use desensitizing gel and lower-concentration bleach to minimize this. Most patients report zero pain with our professional approach.
That’s it. Direct. Answerable. SGE-friendly.
Turning AI Summaries into Appointments
Here’s the real question: appearing in an AI summary doesn’t matter if it doesn’t drive business.
How to ensure your ‘Book Now’ link appears inside the AI snapshot
When Google surfaces an answer from your site, it can include:
- Your business name and rating
- A snippet of your answer
- A direct link to your booking page or phone number
For this to work, you need:
-
A clear, clickable call-to-action on the page your answer comes from. Not buried in paragraph 4. Right near the answer. “Schedule an appointment” or “Call now.”
-
Your Google Business Profile fully optimized with a “Book Online” or “Request a Service” link. This is what appears in the snapshot.
-
Mobile-friendly booking. If someone reads your AI-surfaced answer on their phone and the booking link goes to a clunky desktop form, you’ve lost them.
I watched a plumbing company lose leads for weeks because their answer about frozen pipes was getting surfaced by SGE, but their booking link was broken. They fixed it on Friday. By Monday, appointment requests had doubled.
The role of high-intent Google Business Profile data in feeding the AI engine
Your Google Business Profile isn’t just for local search anymore. It’s AI fuel.
Google’s AI reads your GBP data—your description, services, prices, availability, photos, customer reviews—and uses all of it to decide if you’re relevant to answer a search query. It also uses GBP data to build the snapshot that appears in SGE.
Most local service businesses treat GBP like an afterthought. A generic description, a few photos, maybe outdated hours.
Instead, treat it like your homepage for AI search:
- Description: Write 750 characters that actually describe what you do, not generic corporate speak
- Services: List every specific service you offer with pricing if possible
- Photos: Include team shots, before/afters, tools in action—not just logos
- Posts: Regular updates about seasonal services, tips, special offers
- Reviews: Encourage detailed reviews that explain what problem you solved
A landscaper in Austin rewrote their GBP description from “Full-service landscape design and maintenance” to “We design and maintain drought-resistant landscapes for central