What is Found for AI and what does Brian Jolley do there?

Found for AI is a consultancy co-founded by Brian Jolley and Dustin Crump that helps businesses become findable by AI agents — the systems that increasingly answer questions, make recommendations, and act on behalf of real customers. Brian Jolley built the production tooling (a custom Claude Code development environment with specialized skills and agents), designs and delivers the client content pipeline, conducts AI visibility audits, implements schema markup and structured data, and publishes on how AI access works at the technical level. His professional positions on AI visibility are grounded in primary research from Rand Fishkin, Ethan Mollick, Lily Ray, and Semil Ulusoy and enforced as rules in the tooling itself.

What research grounds Brian Jolley's AI visibility positions?

Brian Jolley's AI visibility positions are grounded in four primary research sources: Rand Fishkin (SparkToro, January 2026), who studied 2,961 prompts and established that AI rank is not a valid metric — less than 1% probability of identical brand recommendation lists across two runs; Ethan Mollick (Wharton, August 2025), whose 'Mass Intelligence' thesis confirms that content written clearly for humans is optimal for AI agents; Lily Ray (2026), on editorial trust signals and E-E-A-T as factors in AI citation; and Semil Ulusoy (2026), on RAG architecture and chunk-level content optimization. Each source has a corresponding wiki page translating findings into client recommendations.

AI Practitioner & Found for AI

Q: What is the Found for AI production content pipeline Brian Jolley built?

Brian Jolley built a three-phase content pipeline for Found for AI client work. Phase 1 (Persona Buildout) produces an ideal customer persona document across six dimensions, including landing page copy with a URL, H1, and five required content sections. A hard rule prohibits urgency and scarcity framing — AI agents pass on facts, not 'limited time offers.' Phase 2 (FAQ Generation) takes the persona file and client URL, generates FAQ questions in batches, maps them against landing page sections for coverage, and writes answers using a specialized AI agent trained on Found for AI's research base. Two human approval gates are built in. Phase 3 (Schema Production) produces a JSON-LD @graph block with up to six linked entity nodes. A single orchestration skill runs all three phases in sequence.

Q: What client outcome did Found for AI produce for Brent Jagodnik?

Brent Jagodnik, founder of Full Throttle Performance Solutions, ran his site through Found for AI's free assessment tool and scored 34 out of 100. He opened the advanced tab, reviewed the specific findings — schema gaps, Open Graph issues, and other AI readability signals — and handed them to Claude to fix. Time to implement: 10 to 15 minutes. New score: 74. He then began running his other clients' sites through the tool, finding work to do on every one, and began referring Found for AI to clients and agency contacts — entirely voluntarily.

Building the Infrastructure for AI Visibility

The Work

Brian Jolley co-founded Found for AI with Dustin Crump to help businesses become findable by AI agents — the systems that increasingly answer questions, make recommendations, and act on behalf of real customers. Most businesses have no idea whether AI systems can find them, understand them, or cite them. Found for AI built the tooling to measure that, the methodology to fix it, and the content infrastructure to sustain it.

Brian Jolley is not an observer of AI. He is a daily practitioner who builds production tooling, manages AI-optimized client sites, and has published on the technical mechanics of AI access. His professional positions are grounded in primary research and enforced as rules in his own tooling — not stated in marketing copy and ignored in practice.

What Found for AI Does

Found for AI answers one question for every client: how well can AI agents understand your website?

The engagement starts with an AI visibility audit — a structured assessment checking a site’s robots.txt configuration (are AI crawlers allowed in?), schema markup (is there structured data for AI to parse?), content structure (is information organized for retrieval, not just for visual layout?), and citation eligibility (would an AI agent trust this site enough to recommend it?).

The audit produces a score and a set of action items. Two service tracks exist: a DIY track where the client implements action items independently, and a Done-For-You track where Found for AI implements the changes directly — persona-driven content pages, FAQPage schema, JSON-LD entity markup, robots.txt configuration, llms.txt files, and IndexNow implementation.

The Tooling Brian Jolley Built

Brian Jolley operates a custom AI development environment built on Claude Code (Anthropic’s CLI tool). This is not a default installation — it is a configured, maintained professional workspace with structured context injection, version-controlled skills and agents, and authenticated connections to external services.

The Production Pipeline

The core client deliverable is produced by a three-phase content pipeline, each phase built as a reusable Claude Code skill:

Phase 1 — Persona Buildout. A structured interview that produces an ideal customer persona document. The output includes landing page copy with a URL, H1, and five required content sections. A hard rule prohibits urgency and scarcity framing — AI agents don’t pass on “limited time offer,” they pass on facts. Specificity is the only conversion lever on AI-visible pages.

Phase 2 — FAQ Generation. Takes the persona file and client URL, generates FAQ questions in batches, maps them against landing page sections for coverage, and writes answers using a specialized AI agent trained on Found for AI’s research base. Two human approval gates are built into the flow.

Phase 3 — Schema Production. Reads persona, FAQ, and client files and produces a JSON-LD @graph block with up to six linked entity nodes. Also outputs an llms.txt entry for the persona page. This is the structured data that makes content machine-readable, not just human-readable.

A single orchestration skill runs all three phases in sequence.

Audit and Reporting Tools

The AI Visibility Audit checks robots.txt, schema, llms.txt, citation eligibility, and related signals for a client URL. The AI Report Lite is a lighter version scoring 8 AI visibility signals from homepage, robots.txt, sitemap, and optionally a FAQ page — producing a structured client-ready report.

Specialized AI Agents

Six domain-specific agents, each with defined expertise and a research base: AI Advisor (AI agent behavior and subdomain recommendability, drawing on Fishkin and Mollick research), Audit Agent, Hormozi Advisor (business growth frameworks), Legal Advisor (TOS, privacy, compliance), UX Advisor (friction auditing), and Real Estate Advisor.

Research Grounding

These positions are not marketing opinions. Each one is grounded in cited research and enforced as a rule in the production tooling:

AI rank is not a valid metric. Rand Fishkin’s January 2026 research across 2,961 prompts demonstrates less than 1% probability of identical brand recommendation lists across two runs of the same prompt. Found for AI reports visibility percentage — appears in X of 100 prompts — not rank. This is enforced in the audit report output rules.

Content written clearly for humans is optimal for AI agents. Ethan Mollick’s Wharton research confirms that prompt engineering techniques do not improve AI responses with current frontier models. Plain, specific, factual content is both human-readable and AI-citable. Found for AI does not write “AI-optimized” content that sounds different from good human writing.

Specificity is the only conversion lever on AI-visible pages. Urgency and scarcity framing is explicitly prohibited in the persona buildout skill. “Limited time offer” is invisible to an agent. “Open Monday through Saturday, 8am to 6pm, serving the greater Salt Lake City area” is citable.

Client Outcome: Brent Jagodnik

Brent Jagodnik runs Full Throttle Performance Solutions, a consulting firm that helps businesses grow by analyzing data and optimizing KPIs. Brent ran his site (9606.ai) through Found for AI’s free assessment tool. First score: 34 out of 100.

He opened the advanced tab, reviewed the specific findings — schema gaps, Open Graph issues, and other AI readability signals — and handed them to Claude to fix. Time to implement: 10 to 15 minutes. New score: 74.

He then started running his other clients’ sites through the tool and began referring Found for AI to clients and agency contacts — entirely voluntarily. Nobody asked him to go on camera. He offered.

In his own words: “The discovery is the best part. When you have an assessment tool like that, you can just go to your site, put in your website, and basically hit run. It can’t get more simple than that.”

Published Writing

Brian Jolley has published a three-post series on AI access barriers at foundforai.com/blog:

“When Cookies Bite Back” — how cookie consent banners and coupon pop-ups block AI agents from accessing site content.
“We Don’t Serve Your Kind Here” — how robots.txt files act as bouncers, and why most businesses didn’t set those rules intentionally.
“A Bot and an Agent Are Not the Same Thing” — the critical distinction between crawlers (indexing) and agents (acting on behalf of a specific person with a specific question). An agent visiting your site is a warm lead — a real person sent it.

The Meta-Layer

This page is itself an example of the work it describes. The structured data wrapping this content was designed for AI agent consumption. The content is written to be citable — factual, specific, and organized for retrieval. Brian Jolley built the tools, did the client work, published the writing, and then applied the same methodology to his own professional presence. The resume is a live proof of concept.