AI visibility sits at the intersection of search, brand, reputation, content, and the public record.
So does my career.
I began my career as a business and technology journalist, where the job was to look past what a company said about itself, examine the evidence, and write the account readers could trust.
Since then I have led brand and marketing across the disciplines AI visibility now touches: search, brand strategy, brand governance, public relations, social media, content operations, and analytics. That has included enterprise brand and marketing communications for a $1.5 billion organization, and content strategy and SEO for a $16 billion, Fortune 500 company.
Along the way:
- Built an enterprise's first brand architecture, identity toolkit, editorial voice, and product-naming guidance.
- Launched an SEO strategy that produced 30% average year-over-year organic traffic growth. By 2025 the organization ranked first in search for more than 5,500 terms.
- Led global media relations and campaigns that earned coverage in 60 Minutes, NPR, The Wall Street Journal, The New York Times, and CNN.
I built AIWorthy to provide independent research and measurement for companies and agencies that want to understand how AI systems find, interpret and recommend brands — and whether that version matches the public record.
That breadth matters because the diagnosis changes the decision. A crawl problem is not a positioning problem. An outdated directory listing is not a reputation problem. A firm that is never retrieved faces a different challenge from one that is retrieved repeatedly and still not recommended. AIWorthy identifies which problem the evidence supports, what the findings do and do not show, and where the evidence stops.
Why the research began with regulated firms
You cannot measure accuracy without something to check it against.
Registered investment advisers file Form ADV with the SEC. Accounting firms performing certain federally funded audits are named in federal records. Those sources exist independently of a firm's marketing, which makes it possible to compare the public record with the AI retelling.
Where no regulator provides a ground truth, the question changes: can AI find the firm when a buyer describes a specific, verifiable service it provides?
That question may be commercially more important than whether AI can summarize a company after being handed its name.
How the work is done
Panels come from a census, not a convenient list.
A directory may be paid, curated or self-selected. Using one to study discoverability can preselect the firms already investing in being found. Wherever the market permits it, AIWorthy starts from a public register and documents the inclusion rules.
Questions are derived, not improvised.
Each firm-specific question begins with a dated public source showing what the firm does. No language model is used to invent the question set.
The method is frozen before collection.
The panel and the questions are fingerprinted before the first response is collected, so the test cannot be quietly altered after the findings appear.
Automation proposes; a human verifies.
Every named firm is checked against the raw response before it enters a finding. A script can identify a candidate. It cannot decide what the evidence means.
Corrections remain visible.
When we make a mistake, we publish a dated amendment rather than silently replacing the record.
The line I do not cross
AIWorthy does not sell optimization, content, remediation, or a retainer designed to improve a result we published.
The measurement is the product.
That is the structure that lets a firm cite the finding, an agency trust us with a client, and a reader evaluate the work without wondering what we hope to sell next.
Based in Apex, North Carolina.
hharreld@aiworthy.ai