From 602 to 2,300 Keywords, and Top 3 on the Map Across Two States.
Zero Ad Spend.
A mortgage lender running branches in two very different markets — a Michigan lakeshore town and downtown Atlanta. We built a page for every loan product in every city they serve, then structured each one to get cited by AI. In months, ranked keywords nearly quadrupled, AI-cited pages went from 7 to 71, and both branches now sit in the top three across their metros.
The Situation — Two Markets, Near-Zero Local Visibility
This mortgage lender runs branches in two markets that could not be more different: Muskegon, a lakeshore town on Michigan’s west coast, and downtown Atlanta, one of the most competitive mortgage markets in the country. In both, they were up against national lenders and rate-comparison directories with enormous domain authority and budgets.
Their reputation and loan products were strong, but they were barely showing up when someone in either metro searched for a mortgage. Organic traffic was small, AI engines did not know they existed, and the map pack was owned by bigger names.
A mortgage lender needs a page for every loan product in every city they serve. FHA loans in Muskegon. VA loans in Grand Haven. Jumbo loans in Atlanta. Without those pages, Google and the AI engines have nothing to show for that exact search — and the lead goes to whoever does have the page.
Step 1 — The Foundation Fix
Before publishing a single new page, we rebuilt the foundation: schema that tells Google and AI exactly what the business is and where it lends, a hub-and-spoke architecture, internal linking, and an answer-first page structure. Most agencies skip this and wonder why rankings never stick. We fix the foundation first, then scale on top of it.
That is what makes pages get indexed in days instead of months, and what makes an AI Overview cite a specific city page instead of ignoring the site.
Step 2 — The Build
Then we built the pages. Every loan product — FHA, VA, USDA, conventional, jumbo, refinance, HELOC, construction, down-payment assistance — for every city across both metros. Each page targets the exact search a borrower makes, and each is structured to be pulled as a source by ChatGPT, Gemini, and Google AI Overviews.
The Numbers — Before vs Now
The keywords did not grow slowly. Organic keywords went from 602 to roughly 2,300, and AI-cited pages went from 7 to 71 — a tenfold jump in how often the AI engines pull this lender as a source. This is Semrush data, not an estimate.
Top 3 on the Map Across Two Metros
The detail that matters most is where those rankings come from. In Muskegon, the branch now holds an average map rank of 2.44, in the top three across 86% of the grid. In Atlanta — a far larger and more competitive market — the branch averages 2.62, in the top three across 80% of the grid. Two very different markets, same system, same result.
Cited by AI, Live
The same pages that rank on the map are the ones AI reads. Ask ChatGPT for a mortgage lender in either metro and the brand comes up by name, with zero ad spend behind it.
What Made the Difference
Two things, in sequence. First the foundation — schema, architecture, internal linking — so every page gets indexed fast and stays ranked. Then the page volume — every product in every city, structured for both organic search and AI citation.
A regional lender now ranks in the top three across two metros in two different states, is cited 71 times by AI engines, and spends zero dollars on ads to do it. Specific pages beat generic directories for specific searches, every time.
Enter your URL and see exactly how many pages the Visibility Engine would build for your branches — every loan product, every city, published in 48 hours.