what to do with my past-client database when rates dip

A dip opened this week. The 30-year fixed eased to roughly 7.12% by October 25, down from 7.40% earlier in the month, and the FOMC decides Wednesday the 28th. You already own the list that benefits most, sitting in your CRM, and the honest problem is that you can't hand-call all of it before rates move back.

Quick answer: To work your database when rates drop, don't blast the whole book. Pull the near-the-money slice you already own, past clients and prequals whose current rate is at least 50 basis points above today's, re-run their quotes, and reach out this week. Then keep the follow-up going after they reply, because the deal is made in the second and third touch, not the first.

How short is this window, and what do I do this week versus this quarter?

The window is days, not months, so treat this week as triage and the quarter as the build. The 30-year fixed closed the week of October 19-23 at 7.25%, a net move of about seven basis points, with the 10-year Treasury back under 5%, per Mortgage Daily. Markets are pricing roughly an 84% chance the Fed holds Wednesday, which means the benchmark could still swing a quarter point either way on the announcement. A borrower closing within 30 days has a real case to lock a re-run quote before the decision rather than gamble on it.

This week you work the narrowest high-value list first: the past clients and prequals who are actually in the money today. This quarter you fix the reason you couldn't work the whole list in time. Those are two different problems, and the ranking guides that tell you to "email the entire aged database" the day rates drop skip the first one entirely.

Which past clients are in the money at 7.1% today?

The candidates are borrowers whose current rate sits at least 50 basis points above today's number. LendingTree's rate guidance puts the rough threshold for a refinance being worth the closing costs at 50 basis points of improvement or more. At a 30-year fixed near 7.1%, that means anyone you closed at roughly 7.6% and up is a live conversation, and your 7.375% files just crossed the line this week. A borrower you closed at 5.75% is not a candidate in a 7% market, and texting them "rates dropped" burns a touchpoint and your credibility.

So the slice is narrow and it is specific. Pull it by current rate, not by "everyone I ever funded." As ezLoanDocs puts it, a borrower at 7.375% becomes a live candidate the moment rates drop meaningfully below that, while the borrower at 5.75% never will. The mistake nearly every rate-drop blast makes, noted by loanofficer.ai, is treating a borrower at 7.5% and one at 7.25% as the same target when they want different things.

Pull the list without doing the math by hand

Filter your CRM and loan origination system (LOS, the system of record for the loan itself) on two fields and you have most of the list in minutes.

Layer in HELOC and PMI-removal candidates if your pricing supports them: borrowers who've built equity since 2023-2024 may win even without a rate story. The output is a working list of a few dozen to a few hundred files, not the whole book. That is the list you work this week.

Who keeps the conversation going after the borrower replies?

Right now, a person does, and that is exactly where the dip leaks. Your automated sequence in Follow Up Boss fires the first text, the borrower replies "what would my payment be?", and the sequence stops, because that is how drip campaigns are built. From there it's on whoever remembers to re-run the quote, send it, and follow up when the borrower goes quiet. In a busy dip week with fifty live replies, that memory fails. Setshape's platform data shows it takes around 17 attempts to convert an opportunity, and the attempts that don't happen are the ones a person was supposed to remember during the exact week the most deals are in play.

This is the leak SalesOS is built to close. It works inside the Claude or ChatGPT account your team already pays for, keeps following up after the first reply in your team's own voice and templates, and keeps going until the borrower acts or is clearly dead, instead of dead-ending the moment they respond. It also nudges the deals that stall waiting on a document after the borrower says yes, so the file you reopened this week doesn't go cold waiting on a pay stub next week. You can see how AI Cofounder works if you want the longer version.

To be plain about what it won't do: it doesn't make rate predictions, it doesn't give borrowers loan or qualification advice, and it follows the texting rules your team and your compliance people set. Your CRM's compliance rules on automated texts still apply, and your compliance team still decides what's allowed. The system works the follow-up; it doesn't override the rules you work under.

Manual blast versus a worked list with follow-up that keeps going

The difference is not effort on day one. It's what happens on day three, when the replies come in and the quiet starts.

Database reactivation during a meaningful rate decline commonly sees 3-8% close rates, versus 1-3% in a flat market, according to workagedleads.com, and the biggest variable is the rate spread, which is precisely why the narrow in-the-money slice beats the blast. Prestyj's benchmarks put past-borrower refi outreach at 6-15% conversion with $1M-$10M+ in recoverable loan volume in a typical book. The volume is real; the question the other guides don't answer is whether anyone keeps working it after the first reply.

Common pitfalls when working the dip

The errors that cost you the window are predictable, and most of them come from treating a short dip like a full campaign.

For a step-by-step version of the segmentation, the rates just dropped, which past clients do I call first guide walks the call order, and how to reactivate past clients when rates dip for a refinance covers the quarter-long version of what you're triaging this week.

FAQ

How do I pull a past-client refinance list without doing the rate math by hand? Filter your CRM and LOS on current interest rate and loan age. Set the rate filter at today's par plus 50 basis points (roughly 7.6% and up at a 7.1% market) and the age filter at two years or more, and the resulting list of a few dozen to a few hundred files is your near-the-money slice. You don't score every file by hand; you let the two fields do the first cut.

What do I say to someone I closed at 7.375% versus someone at 5.75%? The 7.375% borrower gets a re-run quote this week, because today's rate near 7.1% clears the 50-basis-point threshold where a refi starts to pay for its closing costs. The 5.75% borrower gets nothing rate-related, because they aren't a candidate in a 7% market and a "rates dropped" text only teaches them to ignore you. Different current rates are different targets, not one blast.

Do I need to buy another tool to work the dip this week? No. The list lives in the CRM and LOS you already pay for, and SalesOS runs the follow-up inside the Claude or ChatGPT account your team already has, so there's no new app or extra subscription. The question isn't which tool to buy; it's whether anyone keeps working the list after the first reply.

Is reactivation worth it given closing costs? It's worth it only for the borrowers who clear the 50-basis-point threshold, which is why the narrow list beats the blast. LendingTree's guidance is that a refinance generally makes sense when the new rate is at least half a point below the current one, so the files you work are the ones where the savings actually cover the costs. For everyone else, you're spending a touchpoint you can't get back.

If your dip this week is going to leak the same way the last one did, past that first reply and into whoever happens to remember, the fastest way to see what closing that leak looks like on your own database is to book a demo and walk us through your process.