I Audited a Workplace
Culture Consultant.
Search Engines Thought
She Was a Glass Artist.
By Lesli Rose · September 3, 2026 · 7 min read
This consultant has something most of her category fakes: a real method. Five named workshops. A six-step process that ends with pre and post pulse surveys. A homepage argument that takes on Myers-Briggs and DiSC by name and wins. Decades inside a large public-sector institution before she ever sold a workshop.
When I looked at what the open web says about her, the method was nowhere. Google returned galleries and a print store. LinkedIn carried a company name the website never uses. A second live website sold the same workshops under a third name, with the founder bio copied word for word. And underneath the four-item menu on the main site, the sitemap still held twenty-seven pages and a thirteen-category art store from the site's earlier life.
If you run more than one business under your own name, this one is for you.
The Scores
68
Technical SEO
44
On-Page SEO
25
Content
41
Schema
42
AI Discoverability
22
Social SEO
2
Earned Visibility
Technical is the bright spot, which surprises people who assume a site builder means a weak foundation. Earned Visibility at 2 is the story. The owned layer averages 44 and is fixable in weeks. Off the site there is nothing at all for a machine or a buyer to check her against.
Finding #1: Four Names, One Founder
The consulting site used one name. A second consulting site on a different domain used another, with the same workshops and the same founder sentence. LinkedIn and Facebook used a third, a holding-company style LLC. The art business had a fourth. A search for the founder's name returned the art world: gallery pages, a print-on-demand shop, an artist marketplace. A search for the consulting brand plus her name returned nothing from any third party.
An AI assistant asked "who runs workplace culture workshops in this state?" needs one confident entity to name. It reads three names and two sites with identical copy as noise, and the answer goes to a national firm. A buyer who hears her name at a conference and looks her up meets an artist first, then a consultancy called something else. Each hop costs trust before the call is ever booked.
Finding #2: Problem Yes, Process Yes, Price No, Buyer No
I judge a consulting site on four questions. Does it say what problem it solves? Does it show the process? Does it state the price? Does it name the buyer? This one went two for four.
Problem: yes. Disengagement, stated in the reader's words, with a dollar figure and a sharp critique of personality assessments.
Process: yes. Six named steps, five workshops, a twelve-week virtual series, integration modules, measurement at the end. Better than most.
Price: no. No number, no range, no starting-at figure. The page named "Packages" was empty. Every path ended at a discovery call.
Buyer: values only. "Leaders who genuinely care about their people" is a filter, not a target. No sector, no size, no role, no city. The site did not state what city she works from at all.
The irony is that the most credible fact about her, decades of leadership work inside a large public institution, sat on the About page and was never turned into a market. Public sector is the obvious first buyer, and no competitor in her region claims it.
Finding #3: The Consulting Site Was Still an Art Store Underneath
Four pages in the menu. Forty-one URLs in the sitemap. Thirteen art portfolio pages, thirteen store categories of prints, and the most important sales page on the site, the workshops page, living at a slug that literally read "copy of about" and sharing the About page's title and description. Google saw two pages both claiming to be About, and a crawl budget spent on Pop Art and Cartoon before it reached Workshops.
Two typos were baked into the master header: one in the title tag, one in the tagline, repeated on every page. Neither was a ranking factor. Both were the first thing a searcher read.
Finding #4: The AI Summary File Described the Other Brand
The builder platform had auto-generated an llms.txt file and an agent endpoint, which is genuinely ahead of most consultancies. The opening paragraph of that file, the one an AI reads first, described the second brand and its signature experience inside the first brand's file. The one asset built for machines confirmed the confusion instead of resolving it.
Finding #5: Two Audiences, No Bridge
The homepage sold to organizations. The live funnel, two complimentary webinars on mindset and stress, sold to individuals. Nothing told a webinar attendee that the session was a taste of a program their employer could buy, and nothing told a department head that the webinars existed. Both audiences are legitimate. The bridge between them is a paragraph and a link, and it was missing.
What's Actually Working
The argument. A homepage that takes a real position against the assessment industry and explains why. Most sites in this category say nothing anyone could disagree with.
The method. Named workshops, a numbered process, and measurement at the end. That is a case study waiting to be written from data she already collects.
The technical floor. HTTPS, canonical, Organization, LocalBusiness, and Person schema, a clean robots file, and an AI-readable summary file. The builder did its job.
A live funnel. Two scheduled webinars with working registration. The distribution problem is fixable because there is something to distribute.
Does This Look Like Your Business?
If you have ever started a second venture under your own name, search yourself the way a buyer would: your name, then your brand plus your name, then your category plus your city. If the results disagree about who you are, the fix is structural, not creative. One name, one domain, permanent redirects, a named buyer, a price anchor, then the earned citations AI actually trusts.
Frequently Asked Questions
Why does running several brands hurt AI visibility?
AI systems build an entity graph: one name, one description, one set of links that agree. Several names across several domains contradict each other, the system becomes less confident about who the business is, and less confident entities get recommended less. One name on one domain, with permanent redirects, is the fix.
Should a consultant publish pricing?
At minimum a range or a starting-at figure. Buyers and AI assistants both ask what something costs. A pricing page with FAQ schema qualifies callers before they book and gives assistants a fact to cite.
What is an ICP, and why does a values statement not count?
An ideal customer profile names a sector, an organization size, a buyer role, and a geography. "Leaders who care about their people" describes an attitude, not a buyer. Search and AI match businesses to queries by sector and place.
Is Wix bad for SEO and AI discoverability?
Not on the basics. This site had schema, canonical tags, and an auto-generated llms.txt. The ceiling is page weight, security headers, and per-page schema depth. The real problems were positioning and entity clarity, which no platform fixes on its own. Here's the full breakdown.
Discover What AI Systems See When They Crawl Your Website
Technical foundations, entity clarity, AI recall, earned citations, and a clear path from being found to being chosen.
Run Your Visibility Report