Same magnet business. One is a storefront a machine can't cite. The other is scored, structured, and answer-engine ready — built automatically.
01 · The before — a real magnet company, this morning
polymagnet.com is up — it's a real Shopify store selling real magnets. But "up" and "citable" are two different things. We pointed the engine at it this morning. Here is exactly what it found.
The site refused the reader. The engine requested the page and got 429 Too Many Requests — the Shopify layer throttles automated readers. Nothing could be scored. To a crawler or an answer engine, the door is shut.
Zero structured data. The homepage carries no JSON-LD, no FAQ schema, no Organization markup — 309 KB of theme, and nothing a language model can lift and quote. A generic "Home – Polymagnet" title is the strongest signal present.
Effectively uncitable to ChatGPT & Claude. A person asking an AI "what is a programmable magnet?" gets no answer that traces back here. The company exists; in the answer layer, it doesn't. That gap is the entire opportunity.
02 · The live demo — do this on screen, right now
Nothing here is a mockup. Every number comes from a live fetch of the page, the moment you press the button.
Go to automarketingengine.com — the single input box on the home screen is the whole product surface.
Enter wholemagnetics.com and press Analyze. The engine fetches the live page — not a cache, not a saved report.
In seconds it reads the site, identifies the business — niche: magnetics / magnets — and routes it to the right scoring model. No one told it the category.
Overall 71, AI-Search 81, with per-role findings underneath — SEO, Content, Technical, Social, Reach — each traced to signals actually found on the page.
03 · The department — the roles that do the work
Each role reads the site, scores its own domain, and produces work you can approve. Two new roles now push past on-page into off-page authority.
Fetches the live page and grades every signal — the score you saw in the demo.
On-page search and answer-engine optimization — titles, schema, the structure AI reads.
Drafts pages and answers built to be quoted by ChatGPT and Claude, grounded in real facts.
Turns the same facts into channel-ready posts, matched to the brand's voice.
Sequences and newsletters drafted from the site's own material — never sent without approval.
Reads the conversion path and proposes the changes that move a visitor to act.
Tracks each score over time so progress is a number, not an opinion.
Reads the niche and competitors to aim the whole department at what matters.
Backlink and outreach engine — finds where authority should come from and pursues it.
Scores experience, expertise, authority and trust — the signals answer engines weigh most.
04 · The gate — the trust story
The engine drafts; a human decides. Approval is one tap from a phone, and every action can be undone.
A role produces the work — a page, a post, an email — and parks it as a proposal. Live to no one.
The client gets a link. No login, no dashboard to learn — it opens straight to the pending item.
One tap to publish, one tap to reject — from the kitchen table, in ten seconds.
Anything that went live can be pulled back instantly. The client is never locked in.
05 · The rebuild — living proof
These are live today, indexed, and scored by the same engine — a real magnet business rebuilt into pages an answer engine can actually quote.
06 · Pricing — and it takes real payments
The whole department, priced for the size of the operation — and the engine now bills for it.
Get found and get citable. The audit, on-page SEO/AISO, and answer-engine structure — kept current.
The working department. Content, social, and email drafted continuously — everything through the approval gate.
The full engine at scale. Off-page authority, outreach, E-E-A-T, and analytics across a portfolio of sites.
07 · The close
The only question left is who operates it first.