The 30-Day Playbook · A giveaway from Matt Mayo's AI talk 7 pages · US Letter · Print or save as PDF
m·m Mortgage Mayo
A Free Resource
From Matt Mayo's AI Talk

The 30-Day
AI Mortgage Site
Playbook.

A working checklist for solo LOs who want what Matt built — without the trial and error.

This is the playbook from "How a Blog Post Got Me a $750K HELOC Lead in 48 Days." Each item is sized to take 1–3 hours. The whole 30 days runs about three hours per week — less time than most LOs already spend on prospecting that doesn't convert.

Don't try to do everything perfectly. Just start.
Matt Mayo
mortgagemayo.com
30/days
≈ 3 hrs / week
Week 1 ≈ 3 ½ hours of work · 3 tasks

Week 1 — Foundation

Make your existing site readable to AI search engines.

This week
Foundation
Next
Content
Then
Partnerships
Finally
Iterate
Add FAQ schema to your existing site
Time · 1 – 2 hrs

HowMost CMS platforms have plugins. WordPress: "Schema & Structured Data for WP" or "Yoast SEO." Webflow: built-in CMS field. Squarespace: code injection in page settings. If you're on a custom platform, the FAQ schema template (separate resource) shows you the exact JSON-LD code to drop in.

Why it mattersAI engines parse structured data faster than they parse paragraphs. FAQ schema is the most-cited content type by ChatGPT, Claude, and Perplexity.

Write 10 question-format FAQs from real client questions
Time · 1 hr

HowOpen the FAQ schema template for 30 starter questions across categories. Pick 10 that match what your actual clients ask. Don't make them up — use questions you've answered in the last 30 days.

Why it mattersConversational queries are how people talk to AI. Generic FAQ pages don't appear in AI answers. Specific, scenario-based ones do.

Request reviews from your last 30 closed clients
Time · 30 min

HowOpen ChatGPT or Claude. Prompt: "Draft a personal text I can send to a past mortgage client asking for a Google review. Make it warm, not formal. Reference that we closed together but don't get specific. Should feel like a normal text from someone they know." Customize and send to 30 past clients. Aim for 5–10 new reviews.

Why it mattersReviews are your local-pack ranking lever AND an E-E-A-T signal AI engines use. Most LOs have asked zero past clients for reviews.

Week 2 ≈ 4 hours of work · 3 tasks

Week 2 — Content

Build your first piece of hyperlocal content and set up tracking.

Last
Foundation
This week
Content
Next
Partnerships
Finally
Iterate
Build one hyperlocal page for your strongest market
Time · 2 – 3 hrs

HowPick the city or neighborhood where you do the most business. Open Claude or ChatGPT. Prompt: "I'm a mortgage broker in [city, state]. Write a 1,500-word page about getting a mortgage in [city] that includes: local market context, popular loan programs for the area, average home prices and what monthly payments look like at current rates, FAQ section with 8–10 conversational questions, and clear contact CTA. Voice: conversational, expert but approachable. NMLS [your number]." Edit for accuracy. Add real photos if you have them. Publish at /locations/[state]/[city] or similar.

Why it mattersHyperlocal pages are how AI search assigns you to specific markets.

Generate one realtor office market update for your top agent
Time · 1 hr

HowPick your most active referral partner. Use AI to generate a one-page market update with current rates, recent Fed action, jobs data, housing inventory in their market, and a "what realtors should tell clients" section split between buyers and sellers. Brand it to you. Send it to them.

Why it mattersThis single asset opens partnership conversations.

Set up Search Console and GA4 with custom AI traffic channel group
Time · 30 min

HowSearch Console: Add your domain at search.google.com/search-console. Verify ownership. GA4: Make sure tracking is installed. In GA4, go to Admin → Data Display → Channel groups → Create custom group called "AI Traffic" with regex pattern: chatgpt\.com|perplexity\.ai|openai\.com|gemini\.google\.com|copilot\.microsoft\.com|claude\.ai|you\.com|poe\.com

Why it mattersYou can't optimize what you can't measure.

Week 3 ≈ 3 hours of work · 3 tasks

Week 3 — Partnerships

Produce assets that partners can use, and reach out to new partners with personalized AI-drafted messages.

Last
Content
Before
Foundation
This week
Partnerships
Next
Iterate
Build one Nano Banana co-branded flyer for a top agent
Time · 30 min

HowOpen Gemini (also called Nano Banana). Use the prompt template (separate resource). Insert your headshot and a placeholder realtor headshot. Generate. Iterate 2–3 times. Send to the agent.

Why it mattersCo-branded assets compound — her followers see your name, other agents see her sharing it, new partnerships form inbound.

Personalize outreach to 10 new agents using AI research
Time · 1 – 2 hrs

HowPick 10 agents you'd like to work with. For each, prompt Claude or ChatGPT: "Research [agent name] at [brokerage]. What are their recent listings? What's their specialty? Write a brief outreach email referencing their most recent listing, offering a co-branded buyer pre-qual flyer for that listing. CTA: 'Reply yes if useful.' Sign as [Your Name], NMLS [number]." Review each email, edit for accuracy, send.

Why it mattersReply rates on personalized outreach are dramatically higher than blast emails.

Send your monthly market update to existing referral partners
Time · 15 min

HowTake the market update you built in Week 2. Send it to your existing referral partners. Short message: "Quick monthly market update for your sphere. Forward as useful."

Why it mattersExisting partners deserve the same content quality you're producing for new ones.

Week 4 ≈ 1 ¼ hours of work · 3 tasks

Week 4 — Iterate

Look at the data, identify what's working, double down.

Started
Foundation
Then
Content
Then
Partnerships
This week
Iterate
Test your queries in ChatGPT, Claude, Perplexity, Google AI Mode
Time · 30 min

HowOpen each AI engine in incognito. Search for prospect-style queries: "Best mortgage broker in [your city]" / "Mortgage broker in [your city] for [niche]" / "How do I qualify for [program] in [your state]" Note which engines mention you.

Why it mattersThe ones that don't mention you are your priority for next month's content.

Check Search Console for new conversational queries
Time · 15 min

HowGo to Search Console → Performance → Queries. Sort by impressions. Look for full-sentence queries — these are the AI-prompt-shaped queries. Note ones you appear for at position 5+ — those are next month's content opportunities.

Why it mattersThe queries you almost rank for are the cheapest wins on the board.

Decide which pipeline showed traction. Double down.
Time · 30 min

HowLook at the four weeks. What produced the most signal — leads, referrals, or partnerships? Pick the one that's working and pour fuel on it.

Why it mattersMost LOs spread effort evenly across channels that aren't working. Concentration beats diversification when you've found a signal.

Day 30 The state of your engine

End of 30 days — where you should be.

If you ran the playbook in order, you're not chasing leads anymore — you're standing on a small content engine that's starting to produce them. Here's what should be true by day 30.

Foundation,
then fuel.

That's the foundation. Month 2 is about scaling what worked. Month 3 is about compounding. By month 6, the engine should be producing leads, referrals, and partnerships in parallel — all from the same content infrastructure.

P.S. A few things worth saying out loud

A few honest notes.

01

The system, not the timeline, is what matters. If 30 days is too aggressive, take 60. The order of operations is what compounds — don't skip Week 1 to jump to Week 3.

02

You won't see closed loans in 30 days. That's not how this works. You'll see leading indicators: impressions, queries, inbound inquiries, replies. Trust that closes follow.

03

Your competition is not other LOs. It's invisibility. AI search sees your AI-readable content or it doesn't. That's the whole game.

Stay in touch

Questions? Want to share what worked — or didn't?

DM @mortgagemayo on Instagram.

— Matt
I read every message. If something in here clicked, or broke, I'd genuinely like to hear about it. Send me a screenshot of where you got to.
P.S. If you saw the LoanDeck mention at the end of the talk and want a closer look at the loan comparison tool I'm building, here's the waitlist: loandeck.app. Beta users lock in 50% off launch pricing.