mortgage database reactivation checklist for year end

Every October, the marketing calendar tells you to run the second-pass sweep of your database, and every October it dies around day two. The first touch goes out, the lead replies, a borrower needs a callback, and the person who owned the sweep gets pulled into a live file. The names that were supposed to get worked sit there, same as last year.

Quick answer: A mortgage database reactivation sweep works when you pull three lists in priority order (past borrowers, stalled prequals, never-closed leads), give each one a scheduled multi-touch sequence written in your team's voice, route any reply to a human while the non-repliers keep getting touched, and track which names were actually worked instead of trusting a promise. The failure point is never the plan; it's day two, when whoever owns the touches gets busy.

A loan-officer marketing calendar from ezLoanDocs pins October as the year's second database-reactivation sweep, framed around locking in before rates move, with November turning to year-end gratitude and December to reviewing every channel's lead cost and close rate. This year the timing has a live hook: Freddie Mac reported the 30-year fixed at 7.32% on October 22, down from 7.40% the week before, which Bloomberg noted was the first weekly decline in nine weeks. That is enough of a move to reach out about. It is not enough to make a prediction, so don't. The sweep is operator work, not a consumer rate pitch.

Which three lists actually convert in a year-end sweep

Three lists carry a reactivation sweep, and working them in priority order matters more than working all of them at once. Prestyj names the three most lucrative as pre-qualification leads from the last 6-12 months, past borrowers 3-7 years into their loan, and real estate agent partners who haven't referred recently. For a year-end sweep tied to a rate dip, I'd reorder by intent and start with the book you already funded.

Past borrowers, 3-7 years in

Past borrowers are the highest-conviction list because they've already closed with you once, and a rate move puts a slice of them back in refinance, HELOC, or PMI-removal range. Prestyj reports that when rates drop half a point or more, refinance response rates jump to 12-20%. You won't get that from a 0.08-point dip, but the dip is the reason to open the conversation, not the offer itself. Pull everyone who closed between three and seven years ago, sort by rate on file if your CRM carries it, and work the highest rates first.

Stalled prequals and abandoned applications

Prequals and abandoned applications are the second list because the borrower already handed you real information. ezLoanDocs makes the case that pre-approved buyers who sent documents then went quiet are a distinct, worth-reactivating inventory, often a stronger signal than a cold form fill because they did real work. These are the people who got prequalified in spring, didn't find a house, and never got a reason to come back. Year-end is that reason.

Never-closed leads from the last 12 months

Never-closed leads are the third list and the one most teams write off entirely. The reason they're worth touching: ezLoanDocs notes most loan officers only work a lead for the first two or three weeks, then file it away untouched. The lead didn't die. The attention did. A twelve-month pull of inquiries nobody closed is the loosest of the three lists, so it goes last and gets the lightest touch cadence, but it's free inventory sitting in a CRM you already pay for.

Why the sweep dies on day two, and how to keep it alive

The sweep dies because the plan assumes a person will remember every scheduled touch while also closing live loans, and that person can't. This is the gap every ranking guide skips. They tell you to segment, clean, draft, and automate; none of them account for what happens the morning the first replies come in and your best producer disappears into a file.

Two facts explain the leak. First, loanofficer.ai makes the point that reactivation only produces when it's tied to call blocks, sequences, CRM tasks, and manager review; sending messages with no accountable next step is a brand exercise, not a pipeline. Second, Onyx CRM notes that most CRM sequences run only 7-14 days, so prospects who need 30-60 days to be ready get dropped entirely. A year-end sweep needs touches scheduled out past Thanksgiving, which means it has to run whether or not anyone remembers.

The fix is to separate the two jobs. The scheduled touches to non-repliers should run on their own, every one logged. The moment someone replies, that reply should land on a human, because a positive reply is a sales conversation and the rest is still follow-up. This is where the leak gets closed: the system keeps touching the people who didn't answer while your team handles the ones who did.

This is the part SalesOS is built for. It keeps the CRM list current so the sweep isn't working stale data, runs the scheduled touches in your team's own templates and voice, routes any reply to a person, and shows the owner which names were actually worked instead of a promise nobody kept. You can see how AI Cofounder works if you want the wider picture, but for the sweep the job is narrow: run the touches, catch the replies, prove the work happened.

How to build the cadence and route the replies

Each list gets a scheduled multi-touch sequence across text, email, and a call, with the number and spacing of touches matched to how warm the list is. PMR benchmarks a working past-client contact system at four to six scheduled touchpoints per year including a year-end holiday touch, so the sweep is one deliberate burst inside a year-long rhythm, not a one-off blast.

A workable year-end cadence looks like this:

Two rules keep the cadence honest. An open, click, or download is useful marketing data but not a sales opportunity; a positive reply or an appointment request is the threshold that routes into a human conversation. And the cadence should stop instantly the moment the borrower replies or converts, with frequency capped across every campaign so one person never gets hit from the sweep and a separate drip in the same week.

Write it in your team's voice, not a generic script

The sequence should be rebuilt from your own best converted deals, because generic scripts read like generic scripts and your past clients know your voice. The method I use is to pull the actual touches from five deals that closed well, including a couple that were hard, and reverse-engineer the framework from how you actually got those people back to the table. That is more work up front than pasting in a template, and it's the difference between a sweep people reply to and one they mute.

Where compliance fits, and what to decide before day one

Compliance decisions belong to your compliance team, and the sweep has to be configured around the rules they set, not around what the automation can technically do. Opt-outs have to be honored, in practice immediately; the FCC's new TCPA opt-out rules were adopted September 30 and head toward a November effective date, and whoever owns compliance on your team decides how that gets implemented. If you run Follow Up Boss, its texting rules may block automated texts outright, which is exactly the kind of constraint to surface before you schedule a single touch.

Here is the manual sweep against the system sweep, side by side:

The common pitfalls, in the order they usually bite: working stale CRM data so a third of your texts bounce; blasting all three lists on the same day so your team can't handle the replies; running touches so short that the 30-60 day borrowers never get a second look; and, the big one, trusting that the sweep happened because it was assigned. The whole point of tracking worked-versus-promised is that "we reached out to our database" is not a number. The names that got a logged touch, and the ones that replied, are.

If you want the sweep to run this year instead of stalling on day two, book a demo and we'll walk your actual process, pull your three lists, and build the first sequence from your own best deals.

FAQ

When should I run the year-end mortgage reactivation sweep?

Start the October sweep now and schedule the touches out through year-end, because the ezLoanDocs loan-officer calendar pins October as the second database-reactivation pass, and this year's rate dip to 7.32% on October 22 gives a concrete reason to reach out. Keep the cadence running past Thanksgiving so the borrowers who need 30-60 days to surface aren't dropped by a two-week sequence.

Which past clients should I call first when rates dip?

Call past borrowers three to seven years into their loan first, highest rate on file at the top, because they've already closed with you and a rate move puts a slice of them back in refinance or HELOC range. There's a fuller breakdown of which past clients to call first when a dip hits.

How many touches should a reactivation sequence have?

A year-end sweep runs four to six scheduled touches across text, email, and a call over roughly 25 days, with the lighter later touches going only to people who haven't replied. PMR benchmarks a healthy past-client system at four to six touchpoints per year, so the sweep is one deliberate burst inside that yearly rhythm, not a standalone blast.

Does an email open mean a lead is ready to talk?

An email open, click, or download is marketing data, not a sales opportunity; the threshold that should route a name into a human conversation is a positive reply or an appointment request. Treating opens as interest is how teams waste a producer's time on people who aren't actually raising a hand.

How do I know the sweep actually happened?

Track which names got a logged touch and which replied, because "we worked our database" is a claim, not a number. A system that logs every scheduled touch and flags every reply gives the owner a worked-versus-promised view instead of a status-meeting assurance that the list got handled.