The Loan Officer Market Position Report, Built in One Prompt

A market position report is your own closed loan production read back to you as strategy: which markets you actually own, which lender relationships you’re overexposed to, and which real estate agents in your markets aren’t locked up with another loan officer. You can get one in a single prompt by connecting Model Match to Claude — no dashboard to build, no export to reconcile.

Most loan officers know their numbers. Volume, units, average loan size — those are on the monthly report. What that report never tells you is what the numbers mean next to the rest of the market: whether 38% of your book coming from one city is a moat or a ceiling, whether your top lender is a partner or a single point of failure, and which agents around you are writing eight figures of buyer-side business with nobody defending the relationship.

Looking up a loan officer’s volume has been straightforward for years. Reading your own volume as a position — against the lenders, markets, and agents around you — is the part that needed mortgage market intelligence sitting underneath it.

What a loan officer market position report shows

A market position report is short by design. It answers three questions about your own book, and then tells you what to do about them.

1. Where are you concentrated? Your closed loans, sliced by city and market — ranked, with the markets you’ve touched once or twice separated from the ones you actually work.
2. Where is your risk? The same loans sliced by lender. One partner over roughly 45–50% of your volume is a flag, not a compliment.
3. Where is the whitespace? The highest-volume buyer-side agents in your exact markets and price band — scored by how fragmented their referral business already is.

Everything else in the report — production mix, average loan size, market count — is context for those three.

The prompt

Copy this into Claude with the Model Match connector turned on. Replace the items in brackets to make it yours:

The last line matters more than it looks. Without it you get “deepen your agent relationships.” With it you get “Casey Whitman and Grant Calloway, Newport Beach, $208M of fragmented buyer-side volume between them.”

What comes back

Because Claude reads your production live over the MCP connection, the output isn’t a template with your name on it. In a mid-2026 run for a fictional jumbo originator in Beverly Hills, it returned:

• A production snapshot — $311.6M across 224 loans, $1.39M average, 65 markets touched.
• A ranked geographic footprint — Los Angeles at 38% of the book, more than six times the next market, with Newport Beach and Malibu flagged as large-ticket markets barely scratched.
• A lender concentration table — 59.4% of volume through a single lender, plus the two proven partners already handling $50M+ each who could absorb the shift.
• Six named agent targets with buyer-side volume and loan-officer counts, and three explicitly marked leave alone.
• A four-step 90-day plan built from those specific findings.

View the full sample report → — the complete anonymized version, charts, risk flags, and 90-day plan included.

How to find real estate agents who aren’t locked up with another loan officer

Look at how many different loan officers an agent already sends business to, not just how much volume they write. That one number separates a referral partner you can win from one someone else already owns — and it’s the part of the report worth reading twice.

Finding the top-producing real estate agents in your market is easy, and it’s also not very useful on its own. The agent doing $200M in buyer-side volume is being called by everyone. What you actually need to know is whether anyone has that relationship locked up.

Model Match tracks, for each agent, how many different loan officers appear on their buyer-side transactions. That one number changes the whole exercise:

• 25 or more loan officers — the agent spreads business across a crowd. Nobody owns it, nobody’s defending it, and a good conversation can win real share.
• 3 to 5 loan officers — someone already owns that relationship. It might be worth a call, but it’s a knife fight, not an opening.

In the sample report, the top agent by volume works with 36 different loan officers. The second-highest works with 11. Same market, same price point, completely different odds — and you can’t tell them apart from a leaderboard.

That’s the difference between a list of active real estate agents and a list of agents you can actually win.

How much lender concentration is too much?

A common line is 45–50% of your volume through a single lender. Past that, one change to that lender’s guidelines, pricing, or capacity hits most of your pipeline at once — which is why the report flags it rather than congratulating you on it.

This is the section people argue with, so it’s worth being precise about what it is and isn’t. A dominant lender relationship is usually the result of something working — good pricing, an account executive who answers, a process your team knows cold. The report isn’t telling you to break it. It’s telling you what happens if that lender changes guidelines, repricing, or capacity in a quarter where you have no equivalent alternative running.

The useful version of the fix is almost never “go find new partners.” It’s shifting a slice of volume toward the partners already proving themselves at scale in your book, so the backup is warm before you need it. In the sample, two lenders each handling $50M+ made the 59% problem a two-quarter fix rather than a search.

How Claude builds the report from your loan production data

Claude runs a short, specific sequence of Model Match tool calls against your production — nothing in the report is generated from vibes:

1. getOriginator — resolves your NMLS ID to your production: 12-month volume, units, average loan size, career totals.
2. originatorAnalyticsChart — sliced by city, then by lender, then by transaction type and loan type. Four passes, four sections of the report.
3. listAgents — filtered to your core markets and price band, sorted by buyer-side volume, returning the loan-officer count on each agent record.

That’s the whole chain, and it runs in about 15–20 seconds. Every figure in the report traces back to a record you could open yourself.

Connect Model Match to Claude

Setup happens once and takes about a minute:

1. Open your Model Match Connectors settings, then add the Model Match connector in Claude under Settings → Connectors.
2. Sign in with your Model Match account and approve access.
3. Start a chat, confirm the connector is on, and paste the prompt above.

There are no API keys to manage. Your sign-in is secure, every connection asks for your consent, and the data Claude returns is scoped to your account.

For step-by-step screenshots, see the help-desk guide on connecting Model Match to Claude. Prefer a different assistant? The same data works in ChatGPT — here’s how to connect Model Match to ChatGPT.

How often should you run a market position report?

Quarterly. A market position report is most useful as a repeating measurement — run it at the top of each quarter and the interesting part stops being the snapshot and becomes the delta: did the lender share come down, did the second market grow, did the agents you targeted start showing up in your closed loans.

The same connection answers the follow-ups without a new report — which agents in a market you haven’t touched, how a competitor’s book is trending, or how to look up any loan officer’s volume in a city you’re considering. If you’d rather run those searches by hand than by prompt, they’re the same lookups Market Insights does in the app.

FAQs

What is a market position report for a loan officer? A market position report reads a loan officer's own closed loan production back as strategy: which markets and cities their volume actually comes from, how that volume is distributed across lender partners, and which real estate agents in those same markets have referral business that isn't locked up with another originator. It ends with a short list of specific actions rather than a dashboard.

How do I build one with Claude and Model Match? Connect the Model Match MCP connector to Claude, then send one prompt asking for your production broken down by market and by lender, plus the top buyer-side agents in those markets and how many loan officers each works with. Claude reads your loan officer production data live and writes the report — usually in under a minute.

How much lender concentration is too much? A common line is 45–50% of your volume through a single lender. Above that, a change in that lender's guidelines, pricing, or capacity affects most of your pipeline at once. The practical fix is usually shifting a slice of volume toward partners already handling meaningful volume in your book, not sourcing new ones.

How do I tell which real estate agents are worth pursuing? Look at how many different loan officers an agent already works with, alongside their buyer-side volume. An agent spreading business across 25 or more loan officers has no incumbent to displace, so their business is genuinely in play. An agent using only three or four has an established relationship someone else owns.

Do I need my own data loaded into Model Match first? No. The report is built from closed loan production tied to your NMLS ID, which is already part of the Model Match data set — you point Claude at your NMLS ID and it reads from there. Nothing needs to be uploaded or imported first.