Your Borrowers Miss Rate Locks Because Nobody Called Them in Time

The FHFA published a finding in 2024 that should make every branch manager uncomfortable.

As of Q2 2024, the average rate on existing mortgages sits 2.54 percentage points below current market rates.

That gap means borrowers in your pipeline are acutely aware of rate movements.

They are watching.

They are anxious.

And when a favorable rate window opens and nobody from your branch calls them, they notice that too.

The same FHFA analysis estimated that 1.72 million home sales did not happen between Q2 2022 and Q2 2024 because homeowners were locked into low-rate mortgages and refused to move.

Borrowers understand what rates mean to their financial life.

The question is whether your branch communicates with that same urgency, or whether rate lock timing still depends on which LO happened to check their email that morning.

The Mortgage Problem Branch managers running 10 to 20 loan officers face a structural problem that does not get fixed by hiring a better processor or writing a new training manual.

Rate lock timing is a memory-dependent task inside a business that has too many moving parts to rely on memory.

Here is what that looks like in practice.

An LO is working 12 active files.

Rates move favorably on a Tuesday afternoon.

Three of those borrowers hit their target.

The LO means to call all three.

They call two.

The third gets a voicemail at 4:45 PM.

By Wednesday morning, rates have moved back.

That borrower now needs a longer lock term to close safely, which costs more, or they float and take their chances, which costs more if rates move against them.

Multiply this across 15 LOs and you start to see where extension fees come from.

They are not random.

They are the predictable output of a process that was never systematized.

The community research on this is clear.

Borrowers already struggle to understand what a rate lock actually is.

Many believe that sending an email asking to lock a rate constitutes a confirmed lock.

One borrower documented online that they were told a rate verbally over the phone, then asked to text their commitment, and still did not realize the lock was not active until the system confirmed it.

Another borrower sent an email in response to a rate they saw on a lender’s website, only to be told: that’s not how that works, the website is an advertisement.

Loan officers often assume borrowers understand the mechanics.

They do not.

And when communication is delayed or inconsistent, that gap between borrower assumption and lender procedure turns into anxiety, missed windows, and sometimes deals that fall apart entirely.

What Industry Professionals Are Actually Saying

The mortgage community has been talking about this problem for years.

The consensus from practitioners and observers is consistent across multiple points.

First, the standard for confirmation is higher than most LOs practice.

Best practice is to lock the rate while on the phone with the borrower and follow up immediately with a written Loan Estimate. A one-to-seven-hour gap between an email and someone reading it is not acceptable when rates can move materially in that window. A five-minute phone call where verbal confirmation happens in real time is the standard, and written follow-up should arrive the same day.

Second, the education gap is real and costly.

Loan officers who skip or rush the pre-approval and lock-timing conversation are setting up problems downstream.

Borrowers who do not understand what a float-down option is, or what extending a lock will cost them, make worse decisions and feel more anxious throughout the process.

That anxiety generates more inbound calls, more status check requests, and more deals that wobble.

If you want a practical breakdown of how communication breakdowns erode client trust across similar service businesses, the patterns described in this AI client follow-up guide map directly onto what mortgage borrowers experience.

Third, the lack of written confirmation is a recurring complaint.

Even when a rate is verbally confirmed, borrowers often do not receive written documentation until the Loan Estimate arrives.

In the gap between verbal confirmation and written confirmation, if rates move, borrowers start to doubt whether they are actually locked.

That doubt generates support calls that waste everyone’s time.

The CFPB’s plain-language guidance on rate locks is explicit: a rate lock should be documented in writing, including the locked rate, the lock period, and any fees associated with extending or modifying the lock.

Most borrowers have never read that guidance.

Most LOs do not walk them through it proactively.

By The Numbers: Industry Benchmarks

The public data on lock operations is thinner than you might expect.

MBA’s Weekly Rate Survey measures rates on locked loans each week but does not track where in the origination process those locks occur.

Operational lock timing metrics live inside LOS platforms and secondary marketing systems, not public reports.

What the industry data does give us is enough to quantify the cost of doing this badly.

Metric	Typical Range	Source
Loans requiring at least one extension	10 to 25 percent	Secondary marketing benchmarks
Extension cost, 5 to 7 day increment	0.025 to 0.05 points	Aggregator price grids
Extension cost, 10 to 15 day increment	0.05 to 0.125 points	Aggregator price grids
Pull-through improvement with systematic lock management	3 to 10 percentage points	LOS and PPE vendor case studies
Extension frequency reduction with better cycle time management	10 to 30 percent	Hedging consultant benchmarks
Existing mortgage rate gap below current market (Q2 2024)	2.54 percentage points	FHFA Geography of the Lock-In Effect

To make these numbers concrete: a branch closing 25 loans per month at an average loan size of $380,000, with 20 percent of loans requiring a 10-day extension at 0.075 points, is absorbing roughly $14,250 in extension costs every single month.

That is $171,000 per year in margin erosion from a problem that is fundamentally a communication and timing problem, not a market problem.

For a deeper look at how these numbers translate into operational ROI for AI investment, the analysis in this ROI guide for small business walks through the same type of calculation across service industries.

Strategy 1: Systematize Lock Timing So It Does Not

Depend on Individual LO Memory

The root cause of most lock timing failures is that the decision to initiate a lock conversation lives entirely in the head of one loan officer.

There is no trigger, no reminder, no escalation path if that conversation does not happen.

Fixing this starts with defining the decision rules explicitly.

When should a lock conversation happen? Most retail loans should have a lock decision within 24 to 72 hours of application.

If a loan is not locked within that window, someone should know.

Not because the LO is doing something wrong, but because the process requires a checkpoint.

Those checkpoints need to live in a system, not in someone’s memory. A pipeline review meeting is not a system.

It is a meeting. A system runs whether or not the meeting happens, whether or not the LO is having a slow morning, whether or not the branch manager is traveling.

For branches already running an LOS, this often means configuring lock eligibility rules that prevent locks on ineligible products, setting up automated pipeline flags when loans approach lock expiration without a confirmed extension request, and building LO-facing dashboards that surface which files need lock attention today.

The goal is to take the cognitive load of tracking lock status off individual LOs and move it into a visible, shared workflow.

The AI loan processing guide covers how these kinds of workflow triggers work in practice across different loan types.

Strategy 2: Stop Losing

Borrowers to Communication Delays Borrowers who miss favorable rates because nobody called them in time do not always complain.

Sometimes they just go somewhere else.

They find a lender who happened to reach out at the right moment, or they decide the process is too stressful and wait another year.

The communication standard your branch needs to hit is this: when a rate moves into a borrower’s target range, they hear from you within minutes, not hours.

That is not achievable at scale through manual effort.

An LO watching 12 files cannot monitor rate sheets in real time across all of them while also processing documents, talking to underwriters, and returning calls.

The fix is automated rate monitoring with borrower-specific alert triggers.

Each borrower in the pipeline should have a documented target rate on file.

When pricing moves to or through that target, the system sends an alert to both the LO and the borrower.

The LO gets a prompt to call.

The borrower gets a heads-up that their target may be in range and their LO will be in touch.

This is not a new concept in other industries.

The real estate lead follow-up problem described in this data-backed guide follows the same pattern: the gap between when someone is ready to act and when someone from your team actually reaches them is where deals die.

Mortgage rate lock windows are just a faster, higher-stakes version of the same problem.

Beyond rate alerts, the written confirmation gap is fixable with automation.

When a lock is confirmed in the LOS, a system can immediately send the borrower a plain-language confirmation message that includes the locked rate, the lock expiration date, and a brief explanation of what happens if the loan does not close in time.

That single automated touchpoint eliminates most of the anxiety and inbound calls that come from borrowers wondering whether their rate is actually locked.

Strategy 3:

Build the Infrastructure That Replaces Rate Lock Guesswork The absence of an automated alert system is not just a missed convenience.

It is a structural gap that creates measurable financial exposure for your branch every single month. NBER research published in 2024 found that the post-2020 rate environment reduced between-ZIP household mobility by approximately 16 percent, with an economic value of about $49,000 per moving household.

Borrowers are making enormous financial decisions with rate sensitivity at the center.

The branch that reaches them first with the right information wins the deal.

Building the infrastructure means connecting three things that most branches currently run separately: your pricing feed, your pipeline data, and your borrower communication layer. *Pricing feed integration

• means your system knows in real time what rates are available for each loan profile in your pipeline, not what rates were available yesterday when someone pulled a rate sheet. *Pipeline data
• means your system knows which borrowers have a stated target rate, which loans are approaching lock expiration, and which files are at risk of needing an extension because underwriting is running long. *Borrower communication
• means your system can send targeted, timely messages to the right borrower at the right moment, with the right content, without an LO having to manually trigger each one.

When these three layers connect, the workflow looks like this: pricing moves, the system checks which pipeline loans have a target rate that is now in range, it sends an alert to the LO with the borrower’s name and file number, it sends a pre-approved message to the borrower saying their LO will be calling shortly, and it logs the interaction in the CRM.

If the LO does not confirm the lock within a defined window, the branch manager gets an escalation alert.

That is a complete rate lock communication system.

Most branches have none of it and are paying for the gap every month in extension fees and lost deals.

For a look at how this type of workflow automation connects to a broader operating infrastructure, the AI operating system overview explains how these individual automations fit together as a coordinated system rather than a set of disconnected tools.

Implementation Roadmap

Branch managers who want to move from manual to systematized rate lock management do not need to rebuild everything at once.

The sequence matters. *Week 1 to 2: Audit and define. * Document your current lock process in writing.

At what stage does your branch lock loans? Who initiates the lock conversation? What is the process when a borrower has a stated target rate? What happens when a lock is about to expire? Most branches discover they do not have consistent answers to these questions across all their LOs.

That inconsistency is the starting point. *Week 2 to 4: Configure alerts and checkpoints. * Set up pipeline flags in your LOS for loans that are approaching lock expiration without a confirmed extension.

Create a standard intake field for borrower target rate.

Configure daily pipeline reports that surface lock-sensitive files to the branch manager.

These steps do not require AI.

They require discipline and configuration. *Week 4 to 8: Add automated borrower communication. * Deploy automated lock confirmation messages that fire when a lock is recorded in the LOS.

Build rate alert triggers that notify LOs when pricing moves into a borrower’s target range.

Add a borrower-facing pre-lock educational message that goes out at application explaining how rate locks work, what the lock period means, and what an extension costs.

This is where the communication gap starts to close. *Week 8 and beyond: Connect pricing feeds and monitor performance. * Integrate real-time pricing data so alert triggers use live rates rather than manually updated rate sheets.

Track extension frequency by LO and by product.

Measure pull-through rates before and after.

The performance data tells you where the remaining gaps are.

The AI readiness checklist is a practical tool for assessing where your branch is starting from before you begin this sequence.

How RunFrame Approaches This RunFrame deploys

AI systems for mortgage branches that connect these layers without requiring your LOs to change platforms or your operations team to manage another vendor relationship.

The specific workflows RunFrame builds for rate lock management include AI-driven rate monitoring that watches live pricing against each borrower’s target rate in your pipeline and sends alerts to LOs and borrowers when the window opens, automated lock reminder sequences that escalate from LO to branch manager when action is not taken within a defined window, and lock confirmation messaging that fires immediately when a lock is recorded, providing the borrower with written documentation of the locked rate, expiration date, and extension cost structure in plain language.

For branches with investor relationships, RunFrame can also automate the lock request process itself, reducing the time between the decision to lock and the actual lock execution in your pricing system.

This is not a software subscription your team has to configure and maintain.

It is a deployed system built around your specific process, your investor relationships, and your LOS.

The how it works page explains the deployment process in detail, and the mortgage industry page covers the specific workflows most relevant to purchase and refi-focused branches.

If you are not sure where your branch’s biggest exposure is, the AI readiness scorecard takes about five minutes and gives you a specific picture of which workflows are most likely to be leaking margin right now.

The branch managers who move on this first are not technology enthusiasts.

They are operators who looked at their extension fee line item, did the math, and decided the problem was too expensive to ignore. —

Frequently Asked Questions

**How long does

AI deployment take for mortgage companies?*

• Most mortgage-focused AI deployments take 4 to 8 weeks from kickoff to live workflows.

The first automations, such as rate alert triggers and LO reminders, typically go live within the first two weeks.

Full integration with your LOS and investor pricing feeds takes longer depending on your existing stack and how standardized your lock request process already is. *What does AI cost for a mortgage firm?

• Costs vary by scope. A branch-level deployment covering rate monitoring, borrower alerts, and LO reminders typically runs between $1,500 and $4,000 per month depending on loan volume and integrations.

That figure is usually offset quickly when you factor in the extension fees you stop paying. A single 10-day lock extension on a $400,000 loan can cost $200 to $500 in direct price impact alone. *What ROI can mortgage companies expect from AI?

• The clearest ROI in mortgage comes from two sources: reducing lock extension fees and improving pull-through rates.

Industry benchmarks suggest lenders who systematize lock management see 10 to 30 percent fewer extensions and 3 to 10 percentage point improvements in pull-through.

For a branch closing 30 loans per month, even a modest improvement in both metrics can add tens of thousands in annual gain-on-sale. *Do I need technical staff to use AI in my mortgage business?

• No.

The deployments that work best in mortgage are built for branch managers and LOs, not IT departments.

You need someone who understands your lock desk workflow and investor relationships well enough to define the rules.

The technical setup, API connections, and workflow logic are handled by the deployment team.

Your LOs interact with the system through tools they already use: email, text, or their LOS. —

• If you want to know exactly where your branch is losing margin on rate lock failures, start with the AI readiness scorecard at runframe.ai/scorecard.

It takes five minutes and gives you a specific breakdown of which communication gaps are most likely costing you money right now.

Or if you want to talk through your specific LOS setup and pipeline volume, book a discovery call and we can map out what a deployment would actually look like for your branch.