What Google's AI Mode Milestone Means for Local Home-Service Businesses

TL;DR. Google's AI Mode passed one billion monthly users and queries doubled every quarter in its first year, with planning and "where should I" brainstorming queries growing fastest. For local home-service businesses, the homeowner's shortlist now forms inside an AI assistant before any "near me" search happens. The fix is not new: tighten your Google Business Profile, rewrite service pages around the questions customers actually ask, build a steady review and photo cadence, and add schema that matches visible content. Same foundation, executed with more clarity, now covers two surfaces instead of one.

On May 19, 2026, Google's VP of Data Science and UX Research, Shivani Mohan, published a one-year report on AI Mode, Google's conversational AI search experience. The numbers are the kind that change how you think about local search:

• AI Mode has surpassed a billion monthly active users globally.
• AI Mode queries have more than doubled every quarter since the US launch.
• The average AI Mode search is triple the length of a traditional Search query.
• More than one in six US searches now use voice or images, with image searches growing over forty percent month-over-month.
• Planning queries have grown eighty percent faster than AI Mode queries overall in the last six months.
• Brainstorming queries (starting with phrases like "where to," "where should I," and "ideas for") have grown thirty percent faster than queries overall since launch.

Source: "A look at the first year of AI Mode in the US", blog.google, May 19, 2026.

If you run a plumbing, HVAC, electrical, roofing, or other home-service business, none of these stats are about you directly. They describe how the people who hire you are behaving. That is exactly why they matter.

Where do customers actually look first now when they need a home-service business?

A year ago, the path looked like this:

1. Customer's water heater fails.
2. Customer types "water heater repair [city]" into Google.
3. Customer scans the top three Google results plus the map pack, picks one, calls.

Today, an increasing slice of that journey looks like this:

1. Customer's water heater fails.
2. Customer opens AI Mode, ChatGPT, or Gemini, and types a long, conversational message: "my water heater is leaking from the bottom, it is about twelve years old, I am in [neighborhood], do I need to replace it or can it be repaired, and who should I call."
3. The AI returns an answer that explains the likely cause, suggests repair vs replacement criteria, and recommends a small number of local businesses.
4. Customer scans the AI's shortlist, then either calls one of the recommended businesses directly, or runs a quick "[business name] reviews" Google search to verify.

The "near me" query still happens sometimes, but it is no longer always the entry point. The shortlist is being formed upstream, inside an AI assistant. If your business is not in that shortlist, the customer does not see you, and you do not get the call.

Why does AI search matter for local businesses now, not in two years?

The argument that AI search "will matter eventually" was reasonable a year ago. Google just closed that argument with hard numbers:

• A billion monthly users is mainstream scale. It is roughly the size of YouTube's audience a decade ago.
• Doubling every quarter is the kind of growth curve that compounds quickly. A surface used by ten percent of your customers this quarter is used by twenty percent next quarter.
• The query length tripling means people are not just typing keywords, they are explaining their situation and asking for a recommendation. That is the moment a referral happens.

This does not mean classic Google Search is going away. It means a second, earlier search surface has reached scale, and it has different rules.

What does AI Mode reward that classic Search rewards less?

If you boil down what AI Mode behavior looks like under the hood, three patterns stand out for local businesses.

Clarity over coverage

A classic SEO-optimized page often spreads thin across many keywords. AI Mode is built to produce a specific recommendation, so it favors pages that are unambiguous about what you do, where you serve, and why you should be trusted. A page that lists "all plumbing services in the Bay Area" loses to a page that says "we install and repair tankless water heaters in Oakland, Berkeley, and Emeryville, with same-day appointments for emergencies."

Proof, not promises

AI assistants are trained to be cautious about recommendations. They want signals they can lean on:

• A current, complete Google Business Profile.
• A steady stream of recent reviews with substantive content (not just star ratings).
• A real address, real phone number, real photos.
• License numbers, certifications, and credentials where applicable to your trade and state.
• Consistent business information across your site, your profile, your social channels, and your directory listings.

If those signals are missing or contradictory, the AI either skips you or recommends a safer-looking competitor.

Question-shaped content

The article confirms what AEO practitioners have been saying: AI search rewards content that mirrors the question. The fastest-growing query prefixes — "where to," "where should I," "ideas for" — are decision-framing questions, not commercial keywords. Pages that explicitly answer the way a customer asks ("how do I know if I need a new HVAC system?" "what does a slab leak repair actually cost?") perform much better in AI answers than pages built around keyword density.

What stays the same vs what AI Mode actually changes for local SEO?

Most of the foundation does not change. AI Mode reads the same signals classic Search reads — it just weights them differently and is less forgiving of ambiguity.

The classic SEO playbook for home services has not been deprecated. The new AI surface sits on top of it, with sharper penalties for ambiguity and stronger rewards for clarity.

What should I fix this week because of AI Mode?

Four concrete actions, in order of leverage.

Service pages that answer the conversational question

Look at your top three service pages (the ones that drive the most calls). For each, ask: "if a customer described their problem in a paragraph, would this page give them confidence to call us?"

Concretely:

• Add an H2 that mirrors how customers phrase the problem ("My water heater is leaking, what should I do?").
• Add a short, direct answer paragraph (40-80 words) right after the H2.
• Include named neighborhoods, cities, or zip codes where you actually serve.
• State pricing factors honestly (not exact prices, but the variables: unit type, accessibility, permits, age of existing system).
• End the section with a clear next step (a phone number or a booking link), not a generic "contact us."

Photos that work for both humans and image search

Image search growing forty percent month-over-month means a homeowner can now take a photo of a broken fixture and ask AI Mode what it is and who to call. The photos you control — on your GBP, on your service pages, in your reviews — feed that experience.

Practical actions:

• Replace generic stock photos on service pages with real photos of your work, your team, your trucks, your equipment.
• On Google Business Profile, post fresh photos at least monthly. Tag them descriptively (no IMG-2841 style filenames — use names like tankless-water-heater-install-oakland).
• Encourage customers to attach photos to reviews. Reviews with photos are stronger trust signals across every AI surface.
• Include before/after photos on service pages where applicable. AI systems extract these to illustrate answers in image-grounded responses.
• Make sure photos have meaningful alt text describing the work, the equipment, and the location. The alt text is read by both screen readers and AI extractors.
• If you use a service-area business model with no walk-in storefront, photos of vans, branded uniforms, and on-site work are especially important — they substitute for the visual proof a storefront would provide.

Reviews that read like answers, not just star ratings

A four-and-a-half star average tells an AI you are reputable. The content of the reviews tells the AI what you are reputable for. Steer review prompts toward specifics:

• Instead of "leave us a review," ask "could you mention the service we did and the neighborhood you are in?"
• Respond to reviews with substance, naming the specific service and any details that confirm the work. Your responses are part of the review corpus AI reads.
• For negative reviews, respond professionally with specifics — AI systems read response quality as a trust signal.

Schema where it matches visible content

LocalBusiness and Service schema are still worth adding, but only where they match what is visible on the page. AI systems penalize mismatch between structured data and on-page content, because mismatch is a classic spam signal.

What should I NOT do in response to AI search?

A few mistakes to avoid:

• Do not abandon classic SEO. Emergency intent queries still convert at higher rates than any AI-mediated discovery moment. "Near me" + your service + your city remains the highest-revenue keyword family for most home-service businesses.
• Do not chase every AI surface separately. ChatGPT, Claude, Perplexity, Gemini, and AI Mode have differences, but they read most of the same foundational signals. Doing the foundational work well covers most surfaces at once.
• Do not buy "AI SEO" services that promise placements. Like classic SEO, there is no payment that buys a recommendation. Anyone selling "guaranteed AI mentions" is gaming the system in ways that will not last as AI quality filters improve.
• Do not generate AI content at scale to "feed the AI." Networks of AI-written blog posts and listicle articles are a short-term tactic that AI quality filters are already pruning. The durable path is genuine signals: real business data, real reviews, real photos, real service pages.

How do I measure whether AI Mode is recommending my business?

AI visibility is harder to measure than classic rankings because answers vary by user, location, conversation context, and the model's recent training. There is no Search Console for AI Mode, and there will not be one any time soon. The practical, low-cost approach is to build a small repeatable measurement loop.

Step 1: Define your prompt set

Pick ten to twenty prompts that match how your customers actually ask. Aim for a mix of intent types:

• Problem-first prompts ("my water heater is leaking, who should I call in [city]"). These map to the emergency moment.
• Planning prompts ("ideas for upgrading HVAC before summer in [city]"). These map to the +80% planning-query growth in Google's data.
• Decision prompts ("should I repair or replace a 12-year-old furnace"). These are the highest-value moments — the AI is shaping the customer's framework.
• Comparison prompts ("best plumbers in [city] for emergency service"). These produce direct shortlists.
• Disambiguation prompts ("what is the difference between a tankless and tank water heater for a 4-bedroom home"). These build authority over time.

Write the prompts the way a non-technical homeowner would type them, not the way an SEO would. Misspellings, run-on sentences, and missing punctuation are realistic.

Step 2: Run them across multiple AI surfaces

At minimum, weekly:

• Google AI Mode (the surface Google's report is about).
• ChatGPT (consumer market leader).
• Gemini (Google's standalone assistant; overlaps with AI Mode but not identical).
• Perplexity (citation-first; if you appear here, you are extractable).

Add Claude if you serve a more technical or B2B-adjacent customer base.

For each prompt + surface combination, note: did your business appear at all? was it mentioned by name? was it cited with a link? was it recommended as a top option?

Step 3: Track trends, not snapshots

AI outputs fluctuate by design. Any single run can be misleading. Track the same prompts weekly for at least four weeks before drawing conclusions. Pay attention to:

• Does your appearance rate (% of prompts that mention you) go up or down week over week?
• Are you cited with the correct details (services, neighborhoods, credentials)?
• Are competitors gaining or losing ground in the same prompt set?

Step 4: Tie outcomes back to leads

Add an "AI assistant" option to your intake form ("How did you hear about us?"). For new customers, ask the question conversationally. Over a few months, you will see whether the AI-visibility work is producing real lead attribution.

When to automate

Running this manually weekly costs around an hour per cycle for a careful prompt set. That is workable for the first month. After that, it becomes the kind of repetitive work that gets skipped during a busy week — which is when the trend signal breaks.

Automated AI visibility monitoring solves this. Optimizer's AI Search Agent runs a weekly check across ChatGPT, Gemini, Perplexity, and Claude on a prompt set tailored to your services and service area, scores your visibility, and surfaces the specific fixes that move the score most. The output arrives as a simple weekly card in the dashboard, not a raw query log.

Ideas for showing up in AI Mode answers — quick reference checklist

For owners who want a single-page checklist (this section intentionally mirrors the "ideas for…" query prefix Google highlighted):

• Google Business Profile claimed, complete, and updated in the last 30 days
• Primary GBP category is the most specific accurate option (e.g. "Plumber" plus "Drainage service", not just "Contractor")
• At least one fresh photo posted to GBP per month
• Service area lists actual cities/neighborhoods you serve, not a region label
• Top three service pages each have a question-shaped H2 with a 40-80 word direct answer below it
• Top three service pages name specific neighborhoods or zip codes within the body copy
• Top three service pages state pricing factors explicitly (variables, not exact dollar amounts)
• LocalBusiness or Service schema present and matches visible page content
• Steady review cadence (target: 2-3 new Google reviews per month, ongoing)
• Recent reviews respond by name and service ("Thanks for the kind words about the tankless install in Berkeley last week")
• License/certification numbers visible on the site (footer, About, service pages) where required by trade
• Real photos of your team, vans, and completed work — not stock images
• No contradiction between website, GBP, Yelp, and licensing-board listings on hours, services, or address
• At least one "guide-style" reference page on your site (e.g. "repair vs replace your water heater") that competitors do not have
• AI visibility prompt set defined and tested weekly

If you can tick most of these, you are already ahead of the typical local home-service competitor.

The honest summary

Google's report does not say AI search has replaced local search. It says AI search has reached mainstream scale, and the queries it serves are upstream of the commercial keywords your business already ranks for. The shortlist of who a homeowner calls is now formed earlier, inside an AI conversation, before the "near me" search ever happens.

For home-service owners, the message is not "panic about AI." It is "the same foundation, executed with more clarity, now covers two surfaces instead of one." Tighten your Google Business Profile. Rewrite your top service pages around the questions customers actually ask. Build a steady review and photo cadence. Add schema where it matches visible content. Then measure across the AI surfaces with a small repeatable prompt set, not against vanity metrics.

The businesses that do this work in the next three months will be the ones AI assistants confidently recommend when the homeowner is still in their kitchen, before the phone ever rings.

Read next

• Google AI (Gemini) Optimization — run the audit and get the checklist for getting recommended in Gemini and AI Mode.
• AI Discovery Surfaces (AEO/GEO) — overview of how each AI surface differs and where to focus.
• Traditional SEO vs AI SEO (AEO/GEO) — the practical comparison, including what stayed the same and what changed.
• Google Business Profile optimization — the foundation that AI Mode reads first.
• Experience-first local visibility — how to convert AI-mediated visibility into actual calls.

Source

• "A look at the first year of AI Mode in the US" — Shivani Mohan, Vice President, Data Science and UX Research, Google. Published May 19, 2026 on blog.google.