Google has dominated local business discovery for over a decade. If you wanted customers to find your spa, dental clinic, or fitness studio, you optimized for Google. You claimed your Google Business Profile, collected reviews, built local backlinks, and hoped to rank on page one.
That strategy still matters. But the foundation it rests on is cracking.
Reading local reviews and asking for recommendations are different activities. Neither establishes a search-market-share change or proves that an assistant can complete a booking. This guide focuses on testing the latter workflow.
For local service businesses that depend on online bookings -- wellness spas, salons, dental practices, home service providers -- this shift demands immediate attention. This article breaks down where that discovery now happens, what AI-powered discovery means for your business, and exactly what to do about it.
What to Measure
Record how customers discovered your business, whether an assistant submitted the inquiry, and whether the request became a confirmed appointment. Keep those measurements separate.
Test the form and its staff follow-up process outside normal business hours as well as during them.
Use observed customer needs to prioritize the work instead of an unsourced market forecast.
Where AI-Powered Discovery Happens
Customers who skip a search results page find businesses through three categories of AI-powered discovery.
AI Chat Assistants
Consumers increasingly ask ChatGPT, Claude, Perplexity, and Google's own Gemini to find local services. Instead of typing "best dentist near me" into Google and browsing ten results, they say "find me a dentist accepting Blue Cross insurance near Brooklyn" and get a direct recommendation.
The difference is fundamental. Google returns a list of links for humans to evaluate. AI assistants evaluate on the user's behalf and return a recommendation or take action directly.
AI-Powered Search Experiences
Google itself is shifting toward AI-powered search through its AI Overviews (formerly SGE). When a consumer searches for a local service, Google increasingly shows an AI-generated summary at the top of results. This summary often answers the query without requiring clicks to individual business websites.
Bing Copilot and Perplexity operate similarly, providing synthesized answers that reduce the need for traditional website visits.

Autonomous AI Agents
An assistant with suitable tools and permission may take actions as well as answer questions. Its ability to check availability and book depends on the connected scheduling provider and the configured workflow.
MCP can expose implemented operations to a supported client. It does not itself supply availability, payment processing, or booking confirmation.
Why Traditional SEO Cannot Capture This Traffic
If you have invested in SEO for years, your instinct might be to optimize harder. More keywords, better content, stronger backlinks. But the traffic leaving Google is not going to a different search engine that uses the same ranking factors. It is going to AI systems that evaluate businesses entirely differently.
Traditional SEO optimizes for search engine crawlers that rank pages based on keywords, backlinks, domain authority, and user signals. AI-powered discovery evaluates businesses based on structured data, machine-readable service information, programmatic accessibility, and qualification fit.
Here is the critical difference. In traditional search, a consumer sees your listing and decides whether to click. In AI-powered discovery, the AI agent decides whether to recommend you -- and often whether to book you -- before the consumer sees anything. Agents may obtain information from your website, search results, directories, or other sources. Test what your intended client can retrieve rather than assuming your business is invisible.
Consider what happens when someone asks Claude to "find a couples massage in Manhattan within my budget this Saturday." The AI agent needs to discover businesses offering massage services, read structured service catalogs with pricing, check that couples massage is available within that budget, verify Saturday availability, and submit a booking or inquiry.
A visual booking widget or PDF may be difficult for some clients to use. Publish accurate readable service details and consider Schema.org data that matches them. Neither markup nor MCP guarantees that an agent recommends a business or completes a booking.
The Invisibility Problem
Readiness depends on the task and client. Audit your own site and test the intended action.
Many business websites miss the basics an AX Audit checks: they block AI crawlers like GPTBot and ClaudeBot in robots.txt, have no structured data for their services, no llms.txt file, and no MCP endpoint.

A listing or machine-readable page does not guarantee a recommendation. Check that your information is accurate and the request path works.
Compare channels using your own inquiry and booking data rather than assuming that one has less competition.
What AI Agents Need From Your Business
Making your business discoverable by AI agents requires a fundamentally different approach than SEO. AI agents need three things from businesses:
Machine-Readable Discovery
AI agents do not browse websites. They query structured registries and read machine-readable files. Your business needs to be discoverable through: robots.txt configuration that allows AI crawlers (GPTBot, ClaudeBot, Google-Extended), an llms.txt file that describes your business to AI models, Schema.org structured data that makes your services machine-readable, and a .well-known/mcp.json endpoint that tells agents how to connect.
Structured Service Information
AI agents need to understand what you offer in a format they can process. This means your services, pricing tiers, specializations, insurance accepted, availability windows, and location details must exist as structured JSON data -- not as text on a webpage or items in a visual booking widget.
Programmatic Transaction Capability
The agent must be able to take action -- submit a booking, schedule a consultation, or request a quote -- through a clean API call. No form filling, no button clicking, no CAPTCHA solving. MCP (Model Context Protocol) is the emerging standard that provides this capability.
Five Steps to Reach Customers Who Ask AI
Consider these implementation steps for the clients and journeys you intend to support:
Step 1: Audit Your AI Agent Readiness
Run a free AX Audit on your website. It checks 6 dimensions -- Crawlability, Structured Data, Content Quality, Agent Interaction, Discoverability, and Security & Trust -- and gives you a single AX Score. This tells you exactly what to fix.
Step 2: Unblock AI Crawlers
Check your robots.txt file. If it blocks GPTBot, ClaudeBot, Google-Extended, or other AI crawlers, update it immediately. These crawlers need access to understand and recommend your business.
Step 3: Add Structured Data
Add Schema.org JSON-LD markup to your website describing your business type, location, services, and pricing. This is the foundation that AI agents use to understand what you offer.
Step 4: Create an llms.txt File
Add an llms.txt file to your website root. This describes your business in plain text specifically for AI models -- the equivalent of robots.txt but for LLMs.
Step 5: Enable Agent-Ready Booking
Publish a Dashform form and test its MCP endpoint with a supported client. Review the Business Profile, add Services, and enable marketplace participation for listing. Configure any external scheduling handoff separately.
Industry Impact: Who Needs to Act First
The businesses most affected by this shift are those that depend on online discovery and booking.
Wellness and Spa
When consumers ask AI assistants to "find a spa with hot stone massage near me within my budget," only spas with structured service data and MCP endpoints appear. The rest do not exist in this channel.
Dental and Medical
Health-related queries are among the most common AI assistant requests. "Find a dentist accepting Blue Cross near Brooklyn" requires the practice to have insurance information, specializations, and availability in machine-readable format.
Fitness and Personal Training
Class schedules, membership tiers, trial session options, and personal training rates need to be structured data -- not just items on a webpage. AI agents match consumers to fitness options based on budget, schedule, class type, and location.
Real Estate
Buyer agents are increasingly using AI assistants to find real estate professionals. Specializations, service areas, languages spoken, and client types need to be machine-readable.
Home Services
Home management AI agents are emerging that handle maintenance scheduling, emergency repairs, and home improvement projects. Licensed plumbers, electricians, and HVAC providers with structured service data and agent-ready booking capture these automated inquiries.
What Happens If You Wait
The shift toward AI-assisted discovery is happening quickly. Don't wait until most of your customers search this way to make your business visible to it.
Use evidence from your own referral sources and customer interviews to decide how much to invest. Historical analogies do not establish the pace or business effect of agent adoption.
Start with a bounded pilot: verify public information, submit a test request, and confirm which system owns the appointment. Use the observed results to decide the next step.
An AX Audit can identify readiness gaps. Estimate the work to address them, including Dashform configuration, external tool costs, and end-to-end testing.
FAQ
Is Google still important for local businesses?
Yes. Google is still where most consumers read local reviews. The point is not that Google is irrelevant -- it is that relying solely on Google means missing a growing channel. Agent-ready businesses capture traffic from both Google and AI-powered sources.
How quickly is AI-powered discovery growing?
This article does not establish an adoption rate for your customers. Track referral sources and ask customers how they found you before deciding how much to invest.
Do I need to stop doing SEO?
No. SEO continues to drive the majority of local discovery through Google. Agent readiness is an additional layer -- structured data, machine-readable files, and MCP endpoints that AI agents use. Think of it as adding a new storefront for robots while keeping your existing storefront for humans.
What is the cost of becoming agent-ready?
Cost depends on the site changes and external systems required. Review current plan allowances and include development, configuration, and testing effort in your estimate.
Which AI assistants are driving this shift?
ChatGPT (OpenAI), Claude (Anthropic), Perplexity, Google Gemini, and Bing Copilot are the primary consumer-facing AI assistants. Behind them, autonomous AI agents are emerging that book services programmatically without consumer involvement in the search process.
What if my competitors are not agent-ready yet?
Readiness does not make an assistant recommend you by default. Prioritize accurate service information and a tested request path, then measure whether customers use it.







