"AI agents are going to run through every small business in the country. Not a single one of them will know how to build them. Go help them." Mark Cuban, March 2026
Watch the full interview Cuban laid out this thesis in a wide-ranging conversation on TBPN, where he discussed AI agents, startups, and why most small businesses are unprepared. Listen on Spotify or watch on TBPN's YouTube channel.

Cuban is not guessing. He is describing what is already happening. AI agents from ChatGPT, Gemini, Claude, and dozens of vertical-specific tools are starting to book appointments, compare prices, request quotes, and recommend businesses on behalf of real consumers. The question is not whether this shift will happen. It is whether your business will be visible when it does.

We built the tool that answers that question. And after scoring 500,000 businesses, the answer for most of them is: no.

We Built the Tool Cuban Is Talking About

At GradeForAI, we score businesses on AI Agent Preference. AI Agent Preference measures how well AI agents can navigate, understand, and transact with a business. Not how well a business ranks on Google. Not how pretty the website is. How well a machine can actually do business with you.

We measure 4 dimensions:

  1. Agent Accessibility - Can AI agents navigate your site?
  2. Transaction Completeness - Can agents complete bookings, quotes, or payments, and are you connected to agent protocols and booking platform integrations?
  3. Data Reliability - Is your operational data machine-readable, and will extracted data lead to successful operations?
  4. Competitive Position - How do you compare to competitors in your area?

Each dimension gets its own score. Together, they produce an overall AI Agent Preference Score from 0 to 100. Think of it as a credit score for how ready your business is for the agent economy.

When Cuban says "go help them," this is what helping looks like. You cannot fix what you cannot measure. And until now, no one was measuring this.

What the Scores Tell Us

We have now scanned and scored 500,000 businesses across 183+ industries. The results are not subtle. They are bleak.

27 / 100

Average AI Agent Preference Score across 500,000 businesses. That is not agent ready by any standard.

89%
Below Agent Ready
0%
Agent Preferred
~5 / 100
Average Transaction Completeness score (worst dimension)
~28 / 100
Average Agent Accessibility score (best dimension)

Let that sink in. Two out of three businesses scored an F. Not a single business out of the hundreds of thousands we scored earned an A. The best dimension, Agent Accessibility, still averages only about 28 out of 100. And Transaction Completeness, the dimension that matters most when an AI agent tries to actually do something for a consumer, averages about 5 out of 100.

Five. Out of a hundred.

That means almost no business in our database has a way for an AI agent to book an appointment, reserve a table, schedule a consultation, or do anything actionable on behalf of a consumer. The agents can sometimes find these businesses. They almost never can do business with them.

The Gap Between Discovery and Action

The most telling pattern in the data is the massive gap between Agent Accessibility (~28) and Transaction Completeness (~5). Businesses have spent years and millions of dollars making their sites navigable for humans. SEO, Google My Business, social media, paid ads. Some of that work shows up in Agent Accessibility scores.

But almost nobody has done the work to make themselves actionable by machines. Having a beautiful website that ranks on page one of Google does not matter if an AI agent cannot parse your services, check your availability, or book an appointment. The consumer never sees your website. They see whatever their AI agent brings back.

This is the gap Cuban is pointing at. Businesses are optimized for humans with browsers. They are not optimized for agents with APIs.

Where does your business fall? Find out in 60 seconds. Get your free AI Agent Preference Score.

Why This Matters Right Now

You might read this and think you have time. That this is a 2028 problem. Here is why it is not.

AI agents are already booking services. ChatGPT, Gemini, and Claude can already search the web, compare businesses, and present recommendations to users. Early-stage booking integrations are live. The usage numbers are small today. They will not be small for long.

OpenAI's Operator agent booking restaurants and completing tasks autonomously. Source: YouTube

WordPress has signaled plans to add native AI client support in future releases. WordPress powers over 40% of all websites. When it ships with built-in support for AI agent interactions, the infrastructure for agent-mediated commerce goes mainstream overnight. Every WordPress site will have the capability to be agent-friendly. The question is whether business owners will activate it.

Consumer behavior is already shifting. People are getting used to asking AI to do things for them. "Find me a plumber who can come tomorrow." "Book a table for four on Saturday." "Get me three quotes for a fence." Today, the AI gives a list and the human does the work. Tomorrow, the AI just handles it. The businesses that are set up for that interaction win. Everyone else gets skipped.

The competitive window is narrow. This is like SEO in 2005. The businesses that moved first captured organic traffic that took competitors a decade to catch up on. AAO has the same dynamic but on a compressed timeline. AI adoption is moving faster than search adoption ever did. The businesses that optimize in the next 6 to 12 months will have a structural advantage that compounds.

Cuban Is Right. Now What?

Cuban is already walking the talk. As Fortune reported, Cuban bought a Mac Mini specifically to run AI agents for managing his daily workflow, including sorting through 1,000+ emails a day. If a billionaire is restructuring his own operations around agents, the rest of the market will follow.

Cuban's advice was simple: go help them. That is exactly what GradeForAI does.

Step one is measurement. You cannot improve what you do not understand. A free scan on GradeForAI takes 60 seconds, requires no credit card, and gives you your AI Agent Preference Score across all 4 dimensions. You will see exactly where you stand and which dimensions are dragging you down.

For most businesses, the biggest wins will come from three areas:

The bar is extraordinarily low right now. Because 89% of businesses are below Agent Ready, even modest improvements put you ahead of the vast majority of your competitors. A score of 50 would make you an outlier. A score of 70 would make you a category leader.

That will not be true forever. But it is true today.

Check Your Score. 60 Seconds. Free.

See how your business scores across all 4 dimensions of AI Agent Preference. No credit card required.

Get Your Free AI Agent Preference Score

Cuban sees where this is going. So do we. The data from 500,000 businesses confirms it: almost nobody is ready. The ones who move now will not just survive the shift to agent-mediated commerce. They will own it.

Mark Laird
Mark Laird
Founder, GradeForAI. Creator of the AI Agent Preference Score.

Mark coined the term AI Agent Optimization and built GradeForAI to give every service business a clear measure of where they stand as AI agents reshape how consumers find and book services. More about Mark