Google's Search AI agents can now call local businesses on behalf of users. What this means for your website — and how a web developer can get you ready.
In May 2026, at Google I/O, Elizabeth Reid (VP of Search) announced something that caught the attention of local business owners and web developers alike: Google's Search AI agents can now call businesses on behalf of users.
The announcement was part of a broader vision for "Search agents" — AI agents that operate in the background, intelligently reason across information, and take action on your behalf. For select categories like home repair, beauty services, and pet care, Google's agents go beyond searching — they actually pick up the phone and call.
Key fact: This capability is rolling out to all users in the US in summer 2026. While it starts with specific categories, the trend is clear: AI agents will increasingly interact with businesses on behalf of customers.
If you run a service-based business, this changes the game. Your website and online presence are no longer just for human visitors — they're also for AI agents that extract, verify, and relay information to potential customers. A website that's not optimized for AI agents means missed opportunities.
In this article, I explain how AI calling works, what Google's agents look for on your website, and exactly what a web developer can do to prepare your business for this new era.
Here's the scenario: a user in New York needs a plumber for an emergency pipe repair. Instead of manually searching, reading reviews, and calling several plumbers, they tell Google their requirements. Google's AI agent — powered by the Gemini 3.5 Flash model announced at I/O — identifies relevant businesses from Search results, then calls each one.
The agent asks natural-language questions: "Do you offer 24/7 emergency service? What's the average cost for a pipe repair? Are you available this evening?" It then compiles the responses into a synthesized update for the user — letting them compare options without making a single call themselves.
This is not a hypothetical future. Google explicitly stated: "For select categories like home repair, beauty or pet care, you can ask Google to call businesses on your behalf." The agents draw information from:
The AI cross-references these sources. Inconsistencies between your website and your Google Business Profile can cause the agent to skip your listing or relay wrong information to potential customers.
Google's AI agents extract information from websites programmatically. They don't "see" your site like a human — they parse HTML, read structured data, and evaluate page performance. Here's what matters most:
Schema markup is the single most important thing you can add. It tells AI agents in unambiguous machine-readable format what your business is, what you offer, and how to contact you. For local businesses, the key schema types are:
Here's a practical example of LocalBusiness schema for a plumbing company:
{
"@context": "https://schema.org",
"@type": "LocalBusiness",
"name": "City Plumbing Services",
"image": "https://example.com/logo.jpg",
"address": {
"@type": "PostalAddress",
"streetAddress": "123 Main St",
"addressLocality": "Brooklyn",
"addressRegion": "NY",
"postalCode": "11201"
},
"telephone": "+1-212-555-0147",
"openingHours": "Mo-Su 07:00-22:00",
"priceRange": "$$",
"url": "https://example.com",
"aggregateRating": {
"@type": "AggregateRating",
"ratingValue": "4.8",
"reviewCount": "127"
}
}Without this markup, Google's AI agent has to guess your address from text on the page, potentially parsing it wrong. With schema markup, the information is unambiguous.
Your Google Business Profile is the second critical source for AI agents. The agent cross-references your profile with your website. If they don't match — different phone number, different hours, different service list — the agent may have difficulty resolving the conflict.
Common inconsistencies that hurt AI results:
A web developer can audit both your site and your GBP for consistency, updating the website to match your verified GBP data.
AI agents look for phone numbers on your website. The best practice is using the tel: link protocol so both humans and bots can easily identify your contact number:
<a href="tel:+12125550147">(212) 555-0147</a>Place your phone number prominently — in the header, footer, and contact page. Avoid rendering it inside images or JavaScript widgets that AI agents may not parse.
Google's AI agents, like Googlebot, have time budgets for crawling. A slow website (5+ seconds to load) may abort before the agent extracts all the information it needs. Core Web Vitals — especially Largest Contentful Paint (LCP) under 2.5 seconds — are critical.
Speed optimization priorities for AI readiness:
Google's AI agents assess your site's mobile version. Your site must be fully responsive — the mobile version should contain all the same structured data, contact info, and service descriptions as the desktop version. Common mistake: hiding the phone number or address on mobile layouts to save space, which removes it from AI agent parsing entirely.
Here's a practical checklist a web developer can use to audit and prepare your site:
AI Readiness Audit Checklist
tel: link protocolFAQPage schema is particularly valuable for AI calling. When a Google AI agent needs to know "What payment methods do you accept?" or "Do you offer emergency services?", having those answers in structured FAQ format on your website means the AI can extract them directly — no phone call needed.
This doesn't just prepare you for AI calling — it also makes your site eligible for rich results in Google Search, improving visibility for human visitors too.
For businesses that serve specific geographic areas (plumbers, electricians, landscapers), adding areaServed and geo coordinates to your schema helps Google's AI agents determine whether you serve a particular location. This is especially important for the "home repair" category Google specifically mentioned.
"areaServed": [
{
"@type": "City",
"name": "Brooklyn"
},
{
"@type": "City",
"name": "Manhattan"
}
],
"geo": {
"@type": "GeoCoordinates",
"latitude": 40.6782,
"longitude": -73.9442
}AI agents can extract information from image alt text. Ensure your images — especially your logo, storefront, and service photos — have descriptive alt text that includes your business name and service keywords. This adds another signal for the AI to verify your business identity.
Preparing a website for AI agents requires technical expertise across multiple domains: structured data, performance optimization, mobile responsiveness, and business listing management. Most business owners don't have the time or technical knowledge to handle all of this themselves.
As a web developer with 20+ years of experience building production websites, I offer AI readiness audits and upgrades for local businesses. Here's what a typical engagement looks like:
Starting from $300 USD for a landing page audit and fixes. Full AI readiness upgrade for a multi-page business site: typically $1,500–$3,000. Contact me for a free preliminary assessment.
This article builds on several themes I've covered before. For a broader view of how AI is changing search and web development:
Google's AI calling is rolling out now. Businesses that prepare early will have a significant advantage — they'll be the ones AI agents find, extract, and recommend to users.
I offer free initial consultations to assess your website's AI readiness. I'll review your site's structured data, check speed metrics, compare against your GBP, and provide a clear plan of what needs to change. No pressure, no sales pitch — just honest technical advice.
I'm a web developer with 20+ years of experience building and optimizing websites for businesses worldwide. Based in Minsk and working remotely, I help companies prepare for the AI era — one well-structured website at a time.
Tell me about your business website — I'll provide a free AI readiness assessment and preliminary cost estimate.