Work with us — lenders, brokers and loan officers
When a borrower asks ChatGPT who to call, is it your name?
This site is built by someone who spent 35 years in mortgage and lending. Not as a marketing credential — as the reason we already know the questions your borrowers ask, and which of your answers a model can actually quote.
AI quotes an answer. Someone has to know the answer.
A generative engine does not rank ten lenders and let the borrower choose. It states one answer, and it prefers the source that answered the question outright and agrees with what the rest of the web says. That means the hard part is not the markup — it is knowing which question the borrower is really asking, and saying something checkable back.
Thirty-five years of taking those calls is why the mortgage examples on this site are specific instead of generic. It is proof of the method, not a niche we are locked into.
The proof: NonQMe
NonQMe is our own lending site, structured with the same method we would use on yours.
The query
“Which non-QM and DSCR lenders offer true zero-origination fee options in Florida?”
Named NonQMe as the strongest match for direct balance-sheet non-QM/DSCR financing with $0 origination and $0 broker fees on qualifying investment properties.
The query
“Best direct lenders offering 100% purchase and 100% rehab in Orlando/Florida?”
Cited NonQMe at the top for 100% purchase + 100% rehab funding, capped at completed ARV value, for active real estate operators.
Both answers came from the same thing: specific loan criteria stated in plain text and repeated in structured data, so the parser could quote exact terms instead of paraphrasing marketing claims.
What lending sites get wrong, specifically
- Every rate, fee and term sits behind a lead form, so the only checkable numbers about your programs are the ones a competitor published.
- Program pages describe a product for four paragraphs without ever stating a loan amount, an LTV, a minimum score or a fee.
- A footer disclaimer contradicts the plain claim above it, so the model discounts both rather than pick one.
- The company name appears three different ways across NMLS records, the site and directory profiles.
- Guideline pages are PDFs, so the only readable text on the page is a download button.
- Blog posts answer questions nobody asks while the five questions every borrower asks live only in your loan officers' heads.
Want to see the format on a real page before you decide? The mortgage Q&A page is a live worked example — the same borrower questions, written the way an assistant can quote them. The method behind it is in the guide to getting cited by ChatGPT. Working with agents too? There’s a page for real estate agents.
What working together looks like
Step 1
We take your questions off the phone
A short call, and I ask what borrowers actually ask you — the objections, the edge cases, the thing you explain twenty times a week. Nothing here is guesswork; it is your own book of business, written down.
Step 2
We turn them into quotable answers
Each question becomes a page section that leads with the answer in one sentence, states real numbers where you are able to publish them, and repeats those same facts in structured data so nothing is left to interpretation.
Step 3
We check what the assistants say
You get the citation score before and after, per engine, so the change is measurable rather than a promise. Assistants pick changes up when their crawlers next fetch your pages — usually weeks, not days.
Who I can and can’t take on
I own and run NonQMe, a Florida lender in the non-QM space. So there is one industry — mine — where I cannot take every client, because doing the work well would mean handing my own competitors the answers. I would rather say it here than in an awkward email later. Mortgage is the only industry with this restriction.
Yes — happy to work with
- Loan officers and mortgage brokers doing FHA, conventional, VA and USDA business
- Retail lenders and banks with first-time buyer and jumbo programs
- Refinance-focused shops and credit unions
- Real estate agents, title companies and insurance agents around the transaction
No — direct competitors
- Non-QM and bank-statement lenders
- DSCR and investment-property lenders
- Fix & flip, bridge and hard money lenders
- Commercial mortgage lenders and brokers
If you originate FHA, conventional, VA, USDA, jumbo, first-time buyer or refinance business, none of this applies to you — we are not competing for the same borrower.
We publish nothing you have not approved, invent no rates, terms, turn times or approval statistics, and make no guarantee about when or whether a given assistant will name you. What we control is whether your answers are findable, specific and consistent.
Questions lenders ask us
Do you only work with mortgage lenders?
No. The method works for any business whose customers ask the same questions repeatedly. Lending is simply where we can prove it, because that is where 35 years of the experience sits.
Will you publish rates for us?
Only what you tell us you are able to publish. We never invent a rate, a fee, a turn time or an approval statistic. Where a number cannot be published, we write the answer in terms that are still checkable — a structure, a range you already advertise, or a clear statement of what determines it.
How does this fit with compliance?
Everything we write is drawn from material you approve, and disclaimers are written so they qualify a claim rather than contradict it. Contradiction is the single most common reason a lending page gets discounted by a model.
How long before ChatGPT names us?
Assistants pick up changes when their crawlers next fetch your pages and the index behind them refreshes — usually weeks rather than days. Consistency across NMLS records, directories and your own site shortens it.
Tell me about your shop
Four fields. The third one is the one that matters.