Local Business Agent Readiness: The $6.2B Opportunity Nobody Is Chasing
There are 33 million small businesses in the United States. Zero of them have MCP servers. That gap is worth an estimated $6.2 billion a year in infrastructure, and not a single local business in your city is racing to fill it. The first plumber, dentist, and salon in each zip code to go agent-ready claims every agent-driven lead for the next 12 to 18 months. This is the early-web moment for local commerce.
33 Million Businesses, Zero MCP Servers
The Small Business Administration puts the count of US small businesses at 33 million. AgentHermes has scanned 500 across 27 verticals. The distribution is brutal: one Gold, 52 Silver, 249 Bronze, and 199 scored below Bronze. When we filter for local businesses specifically — restaurants, trades, clinics, salons, gyms — the below-Bronze share climbs to nearly 90%.
Developer tools dominate the top of the leaderboard for one reason: they ship APIs, docs, and structured data as a core part of their product. Resend (75), Vercel (70), Stripe (68) are agent-ready because their customers are developers. Local businesses have no such pressure — until the pressure becomes agents making purchase decisions on behalf of their users.
That pressure is here. Claude and ChatGPT both ship MCP connectors. Users already ask agents to book dinner, find an HVAC tech, or schedule a dentist. The agent does its best with web search, reads mangled HTML, and hands back a phone number the user will not call. The business that made itself callable — really callable, with a book_appointment tool — wins the lead.
Why Local Businesses Score ARL-0 Dark
ARL-0 Dark means completely invisible to AI agents. No discovery files, no structured data, no API surface area. Here are the five patterns we see again and again when scanning local businesses — any one of them caps the score in the bottom quartile.
Facebook-only presence
Typical result: ARL-0 DarkThe business lives inside a platform that blocks AI crawlers. No domain, no structured data, no way for an agent to learn hours or services.
Phone-only booking
Typical result: ARL-0 DarkThe only call-to-action is "call us." Agents cannot place voice calls on behalf of users without permissions, so they route the user to a competitor with an online form.
No website at all
Typical result: ARL-0 DarkGoogle Business Profile plus Yelp plus Instagram is not a substitute for a canonical site. Without a root domain there is no place to host llms.txt, agent-card.json, or an MCP endpoint.
PDF menus and price lists
Typical result: ARL-1 Indexed (weak)A PDF is opaque to agents. Prices inside a PDF cannot be parsed reliably, so the agent tells the user to "check the website" — which kills the booking.
Booking locked behind a portal
Typical result: ARL-1 IndexedThird-party booking sites that require login and do not expose an API cut the agent off at the knees. Calendly, Square Appointments, and Fresha are closed silos to most agents.
These failure modes are not a judgment on the businesses. They made reasonable choices for the web of 2015. A Facebook page was free, phone booking worked, and PDFs were good enough for a print-era menu. The agent economy changes the math. A business that is not addressable by an agent is not findable by the agent’s user either.
What Agent-Ready Looks Like for a Local Business
You do not need a venture-backed tech stack. Six ingredients separate ARL-0 Dark from ARL-3 Actionable. Each one can be shipped in a day. Together they are the checklist for the first-mover in your category.
Canonical domain with TLS
yourbusiness.com on HTTPS. No TLS caps your score at 39 — instant Bronze-or-below.
Structured hours and location
Schema.org LocalBusiness JSON-LD in your homepage head. Agents parse it directly, no scraping.
agent-card.json at root
A2A v0.3 descriptor declaring your skills — check_availability, get_quote, book_appointment.
llms.txt with the essentials
Your overview, service list, booking URL, pricing page, and agent-card.json link in one fetch.
Public pricing or quote endpoint
Either structured prices on a /pricing page or an API that returns a quote for a given job. No "call for a quote" dead ends.
Booking via API or MCP tool
book_appointment(service, time, contact) returns a confirmation. This is what turns an agent conversation into a paying customer.
Real example: a restaurant with TLS, a Schema.org LocalBusiness block, an llms.txt that lists hours and menu URLs, an agent-card.json declaring check_availability and book_reservation, and a Square or OpenTable booking API connection. That stack pushes the business from ARL-0 Dark to ARL-3 Actionable. Cost: a Square account and a weekend.
The First-Mover Window in Every Zip Code
When an agent tries to book a service in a zip code and finds only one agent-ready business, it routes every relevant user to that business. This is not a preference weight — it is a hard routing decision. An agent cannot book with a Facebook page. It can book with a Square Appointments API. There is no tie to break.
In the 500 businesses we have scanned, we have yet to find a local zip code with two agent-ready plumbers. Most have zero. The first to move captures the full share of agent-driven demand for the next 12 to 18 months before a second entrant catches up. The dynamics look a lot like Google in 2003 — early SEO investment compounded for a decade.
For consultants, digital agencies, and local marketers this is the highest-leverage product of the next three years. An agency that helps 50 dentists in a region become agent-ready before any other agency figures out the playbook owns that category. The playbook is public, the tools are free, the customers have never heard of it.
Restaurants
First in your city to ship a menu tool plus reservation API captures every "find me a quiet table for two at 7" request.
Home services
Plumbers, HVAC, electricians, locksmiths. Emergencies are priced high and time-critical. The first agent-ready shop wins urgent jobs.
Medical and dental
High booking intent, cash-pay available, appointment APIs already common. Zocdoc-adjacent but cheaper and direct.
Personal services
Salons, barbers, spas, trainers. Recurring customers, known inventory of time slots, mature booking stacks. Low-hanging fruit.
Start by running an Agent Readiness scan on three competitors in your zip code. Most will score below 40. The gap you see in that report is the size of the opportunity you can close in a weekend with free tools and a Square or WooCommerce connector.
How to Become Agent-Ready Without Rebuilding Your Site
You do not need to rip out your WordPress site or migrate away from Wix. Agent readiness is a layer on top of whatever web presence you already have. Here is the minimum viable path.
Run a free AgentHermes scan
Get your current score, identify every missing file, and see a list of competitors in your vertical sorted by score. Most local businesses learn they are ARL-0 Dark in 60 seconds.
Add Schema.org LocalBusiness JSON-LD
Paste a script tag into your homepage head with your name, hours, address, phone, and service catalog. Agents read this directly without scraping visible HTML.
Ship agent-card.json and llms.txt
AgentHermes auto-generates both from your business profile. Upload to the root of your domain or use our /connect wizard to host them for you.
Connect a booking backend
WooCommerce, Square Appointments, Calendly, or a vertical-specific tool. AgentHermes ships adapters that expose your calendar as an MCP tool so agents can actually book.
Publish to the registry
Your agent card gets a listing in the public AgentHermes registry so any agent, not just ones routed through us, can discover your business.
This path takes a weekend to execute manually, or about five minutes through our vertical builder. Either way, you exit ARL-0 Dark and enter the tier where agents can actually interact with your business. See the full $6.2B gap analysis and the restaurant-specific playbook for more depth.
Frequently Asked Questions
Why do local businesses score so low on Agent Readiness?
Most local businesses operate through closed platforms — Facebook pages, Yelp listings, third-party booking tools — with no canonical domain or machine-readable data. Our scans of 500 businesses show 199 below Bronze (score under 40), and the vast majority of those are local service businesses. Without a domain that serves llms.txt, agent-card.json, and structured Schema.org data, an agent has nothing to work with.
Do I need to rebuild my website to become agent-ready?
No. You can layer agent readiness on top of almost any existing site. Start with three files at the root of your current domain — llms.txt, agent-card.json, and a robots.txt that allows AI crawlers. Add Schema.org LocalBusiness JSON-LD to your homepage. Plug a booking API through a WooCommerce, Square, or Calendly integration. You can reach ARL-2 Indexable in a weekend without touching your design.
What is the first-mover advantage for local businesses?
When an AI agent is asked "find me a plumber in 78704," it will pick the handful of businesses it can actually interact with. Today that is zero to one in most zip codes. The first plumber, dentist, or salon in a zip code to publish an agent-ready site captures every agent-driven lead for that category until a second entrant catches up. That is a 6 to 18 month window where you own the channel.
How much does this cost for a single-location business?
The underlying infrastructure is free — llms.txt, agent-card.json, Schema.org markup, and a robots.txt update. AgentHermes auto-generates all four in the Connect wizard. The only costs that may apply are the optional managed MCP server for dynamic tools and whatever e-commerce or booking backend you use. A single-location business can become ARL-2 Indexable for under 100 dollars a month, fully hosted.
Which local verticals benefit first?
High-frequency, high-intent verticals win early. Restaurants (reservations), home services (plumbers, HVAC, electricians), medical and dental (appointments), salons and barbers (booking), auto repair (service scheduling), and fitness studios (class signup). Any vertical where a user tells an agent "book me one" benefits the moment a single competitor in their area becomes agent-ready.
Claim your zip code
Run a free Agent Readiness scan, see your score plus your top three competitors, and generate every agent-ready file you need in the same session.