The traditional sales funnel is dying, and most business owners are still trying to decorate the corpse. For the last twenty years, the 'Path to Purchase' for a local service—whether a plumber in London or a real estate agent in New York—followed a predictable rhythm: search, click, browse, convert. We spent billions on 'eye-catching' websites and high-converting landing pages. But we are entering the era of the Zero-Click Customer, a shift that makes AI for small business less about 'using' tools and more about becoming 'legible' to them.
In this new reality, the customer isn't the person holding the phone. The customer is the AI agent acting on their behalf. When someone says, 'Find me the most reliable cleaning service within five miles that can come this Thursday and handles pet stains,' they aren't going to look at your beautiful hero image or your 'About Us' page. Their AI agent will. If that agent can't 'read' your business logic instantly, you don't exist.
I’ve spent the last year watching these patterns emerge across thousands of businesses. The gap between businesses that are 'human-ready' and those that are 'agent-ready' is widening. If you’re still paying for traditional website design focused purely on aesthetics, you might be investing in a billboard that the most important buyers—AI agents—simply cannot see.
The Rise of the Machine Interview
💡 Want Penny to analyse your business? She maps which roles AI can replace and builds a phased plan. Start your free trial →
We are moving from the era of 'Search' to the era of 'Selection.' In the old world, Google gave a user ten blue links, and the user interviewed the businesses by clicking on them. In the new world, the AI agent performs what I call The Machine Interview.
An LLM (Large Language Model) doesn't 'look' at your website; it ingests your data. It looks for structured proof of capability, availability, and reliability. It parses your Schema markup, your API endpoints, and your crawlable reviews. It asks: Is this business real? Can they do the job? Are they available? Do they have a track record of solving this specific problem?
If your website is a 'black box' of pretty pictures and vague text like 'We offer world-class service,' you will fail the Machine Interview. The agent will move on to a competitor whose data is structured, transparent, and instantly verifiable. This is the new reality of property marketing and other high-intent local services: if the agent can't verify your credentials in milliseconds, the human never hears your name.
The LLM Legibility Score
I’ve started measuring businesses by what I call their LLM Legibility Score. This isn't about SEO rankings. It’s about how easily an AI can synthesize your business into a recommendation.
A high Legibility Score requires three things:
- Structured Availability: Not a 'Contact Us' form, but a machine-readable calendar or API.
- Granular Capability Data: Not 'Cleaning Services,' but 'End-of-tenancy steam cleaning for 3-bedroom properties including oven degreasing.'
- Third-Party Verifiability: Reviews and certifications that aren't just text on a page, but are linked to trusted, crawlable databases.
Many businesses are currently paying what I call The Agency Tax—they are paying marketing agencies thousands a month to optimize for keywords that humans are stoping using, while their actual digital infrastructure is a mess for AI. For example, in the cleaning industry, I see companies spending £2,000/month on Google Ads, while their booking system is a manual 'call for a quote' process. An AI agent cannot 'call for a quote.' It will prioritize the business with an instant-price API every single time.
From Websites to Data Repositories
For two decades, your website was your storefront. In the next decade, your website is just a 'view' of your database. The real value lies in the data itself.
Think about how we use Uber or Airbnb. We don't visit the individual 'websites' of the drivers or hosts. We interact with a central layer of data. AI agents are becoming that layer for the entire local economy. They are the 'meta-platform' that sits above individual businesses.
If you want to win the Zero-Click Customer, you need to stop thinking about 'web design' and start thinking about 'data distribution.' This means:
- Prioritizing Schema.org: This is the language machines speak. If you aren't using deep, nested Schema for every service, price point, and employee bio, you are whisper-quiet in a room full of AI ears.
- Open APIs: Even for a small local business, having a way for external systems to query your pricing and availability is becoming a competitive moat.
- Niche Content Depth: AI models are trained on the web. If you have written the most detailed, structured guide on 'How to fix Victorian plumbing in North London,' the LLM 'knows' you are the expert. When a user asks for a plumber who understands Victorian pipes, the LLM isn't just searching for keywords; it's recalling its 'training' that identifies you as the authority.
The 90/10 Rule of Modern Presence
I advocate for the 90/10 Rule: 90% of your digital effort should go toward making your business data-rich and machine-readable. The remaining 10% is the 'human layer'—the brand aesthetics that close the deal once the AI agent has already shortlisted you.
Most businesses have this backward. They spend 90% of their budget on a 'look' and 10% (or 0%) on their data structure. They are building beautiful stores in a city where every customer is blind and relies on a robotic guide.
The Death of the 'Lead' and the Birth of the 'Booking'
In the Zero-Click world, the 'Lead' is a failing metric. A lead implies a human who needs to be sold to. A 'Booking' is a machine-negotiated transaction.
If you are a local service provider, your goal should be to remove every friction point that prevents an AI agent from hiring you. If an agent has to wait for a human to 'get back to them' with a quote, that agent will find another provider who provides a quote via an automated logic engine. This is why AI for small business is fundamentally an operational transformation, not just a marketing one.
How to Start Improving Your Legibility Today
You don't need a massive tech budget to survive this shift. You need a change in perspective.
- Audit your 'Machine-Readability': Copy your website text and paste it into a tool like Claude or ChatGPT. Ask it: 'Based purely on this text, what are the exact prices, availability, and specific service areas for this business? Can you book them right now?' If the AI says 'No,' you have a legibility problem.
- Kill the PDF: If your pricing or service lists are in PDFs, they are invisible to many real-time agents. Move everything to structured HTML and Schema.
- Invest in Structured Reviews: Encourage customers to leave reviews on platforms that support structured data (like Google Maps or industry-specific portals) rather than just pasting testimonials on your site as 'unstructured' text.
The Opportunity for the Early Movers
The businesses that adapt to the Zero-Click Customer first will experience a massive drop in customer acquisition costs. Why? Because while their competitors are still fighting the 'SEO wars' and paying for expensive clicks, the early movers will be quietly 'hired' by the AI agents for free, simply because they were the only ones the agents could actually understand.
I’ve worked with businesses that have shifted their entire marketing spend from 'ads' to 'data infrastructure,' and the results are consistently leaner, more efficient operations. They aren't chasing customers; they are positioning themselves to be found by the machines that customers now trust.
Your website isn't for people anymore. It's for the algorithms that serve them. Is your business speaking their language?