Build a Mortgage Branch That Can Grow

The meeting everyone was looking forward to

Morgan’s branch had funded $48 million last year. A respected producer, Ellis, was considering coming over with two colleagues. Morgan’s manager wanted to know if they could start the following month. The obvious answer was yes. The branch had been talking about growth for a year.

Then Morgan asked operations lead Priya which processor would own the new group’s first ten files. Priya turned her laptop around. Two existing processors were covering a leave, and a third was training a new hire. The branch could add people to the roster faster than it could give them a dependable place to work.

At the Monday huddle. Morgan asked, “If three producers start together, what are we asking the current team to stop doing?” Priya said she could train the hires or supervise the leave coverage during the afternoon, but both tasks required the same uninterrupted time. They marked each conflict on one calendar. The bottleneck was two hours a day of an experienced person’s attention. Buying more software or issuing a bigger production goal would leave that conflict in place.

The case in this book is fictional. Put yourself in Morgan’s chair. You want Ellis to join, and you’ve just discovered a reason to slow down. If you delay, Ellis might lose interest. If you promise a start the team cannot support, you may spend the next six months fixing a problem you created.

Write a sentence describing the kind of growth your branch can absorb. Include what must remain true for the people already there. Morgan’s first version was simple. “Add production without making our existing producers wonder who’s working their files.” That gave Priya a reason to join the recruiting discussion early.

Try this with your managers. Invite production, operations and finance to write their own definition of a successful start before sharing answers. Production may say “first submissions,” operations may say “complete files entering the right queue,” and finance may ask when contribution covers support. Turn those into separate milestones. If everyone uses a single word such as “ready,” ask for a specific event that proves it. Agree on who can recognize that event without asking Morgan to interpret it later.

Figure out what the production number contains

Ellis reported $18 million last year. The two colleagues reported $7 million and $5 million. A quick spreadsheet showed $30 million. At the next meeting, Morgan asked how the numbers were attributed. The $18 million included $4 million credited to one colleague on shared transactions. The second colleague’s $5 million was separate. The comparable, deduplicated starting figure was $26 million.

Even that number described last year’s funded business. Morgan had to learn how much depended on a branch-generated lead source, a product specialist, or a referral partner who would need to make a fresh decision.

Reconcile the deal rows. Ellis’s $18 million already included the shared $4 million, so the spreadsheet should show $18 million plus $3 million of the first colleague’s independent production plus the second colleague’s $5 million. Morgan kept the shared $4 million visible as one row rather than erasing it. Later, when the team asked whether the referring partner would continue to work with both people, there was a place to record the assumption without counting the loans twice again.

Use one time window and one definition of funded volume. Ask each producer for aggregate figures they’re authorized to share. Mark shared credit as shared. Identify the portion that depends on a channel or team capability you would need to replace.

Morgan used $26 million as historical context and built three annualized planning cases at $13 million, $18 million and $23 million. Those cases were assumptions, with no guarantee of transfer. The branch could now ask which assumptions had to hold for each case. Run the exercise with your own numbers and circle the line you would be most reluctant to defend in front of operations.

Stress one source. Suppose $6 million of the deduplicated history came through a referral arrangement the receiving branch cannot replicate. The remaining $20 million is still historical volume. Try an explicit planning assumption such as retaining three quarters of that amount, or $15 million annualized. Ask who can check the underlying dependence, which parts are unknown, and what new activity could help replace the gap. Keep that new activity outside your base case until the team can describe how it would happen.

Find the actual constraint

Priya ignored the branch’s annual volume target for a minute. She pulled up the past four weeks and counted active files by stage. The bottleneck was getting purchase files through the final conditions window. Average workload looked manageable. Thursday and Friday were where the problems showed up.

They used a rough planning estimate. At an assumed $300,000 average funded loan, $18 million would represent about 60 fundings a year, or five in an average month. Priya expected additional files in process that wouldn’t fund that month. She also knew the group’s files might bunch up around the same closings. An average could hide both problems.

A week on Priya’s board. During a busy week, 11 files arrived at final conditions and the team cleared nine. Two carried forward. If that repeats for three weeks, the queue grows by six, even before adding anyone. The issue becomes visible in unanswered updates and deadline escalations well before it appears in quarterly volume. Priya asked the team to bring the oldest waiting file to the huddle and explain what would move it. That was the beginning of a capacity conversation.

Start with your last eight weeks. Find where a second person gets pulled in to rescue the same task. Count arrivals and completed work at that stage. Ask the person doing the work what happens when one teammate is out. You’ll usually get a more useful answer than “we can handle another five loans.”

Priya offered two starts initially if there was a backup for purchase escalations. The third could come after the new processor could handle that queue independently. Morgan turned the constraint into a condition Ellis could evaluate. It was easier to discuss a real capacity limit than explain a missed promise after a start date.

Use a practical capacity gate. Set a threshold your team can observe, such as confirming ownership on every urgent file during two consecutive busy weeks. Give the backup a turn while the primary owner is available to coach. Morgan would consider a delayed third start if two purchases in the same week waited for an owner. Write your own warning signal and ask operations whether it would give enough time to act. A threshold that only appears after a missed closing is too late.

Compare the economics under pressure

Morgan’s first model made the hire look easy. At an illustrative contribution of $1,800 per funded loan after the variable costs included in the branch model, 60 loans would contribute $108,000. Added fixed support would cost $72,000 annually. That left $36,000 before any additional overhead or transition costs.

Priya asked what happened at 40 loans. At the same contribution per file, the $72,000 covered the added support exactly. If contribution fell to $1,500 per file, the branch would need 48 loans to cover that support. The team now had two assumptions to verify before treating the upside as available cash.

Budget the timing. Morgan’s $72,000 in annual support was $6,000 per month. In the fictional ramp, the first two months yielded two and three funded loans. At $1,800 per loan, the combined $9,000 in contribution fell $3,000 short of the $12,000 support cost before transition expenses. A positive annual projection could still require a bridge. Morgan asked finance what cash reserve would cover a slower start and how long leadership was willing to keep staffing ahead of the pipeline.

Build the model in units and dollars. Keep volume assumptions separate from contribution per file. Confirm the definitions with the finance owner. List any added costs outside that contribution figure, including the cost of overlapping support during onboarding, and keep one-time and recurring costs in separate rows.

Ask what happens if you have to pay for the support a quarter before the production appears. Morgan’s decision was easier once the branch knew it could afford a ramp. If a plan only works when every person starts on time and every file funds quickly, write down which expenses you can delay and what that would do to service.

Do the sensitivity aloud. Try 36 annual fundings at $1,500 contribution. That yields $54,000 against $72,000 of added support, an $18,000 gap before additional costs. Ask which assumption you can influence. You may be able to phase a hire, change how duties are assigned, or wait until the first group is ready. Record the trade-off each action creates. Cutting a service owner to make the spreadsheet balance may leave you selling a support model the branch can’t deliver.

Recruit the group without recruiting past the people

Ellis knew all three producers well and spoke confidently about what they wanted. When Morgan talked with the second producer individually, a different story emerged. Ellis cared about growth support. The colleague wanted predictable hours and a clearer handoff to operations. The third person wanted time to understand the product mix before discussing a date.

Morgan’s opener changed. “Ellis told me what matters to him. I’d like to hear what you would need to see before you’d even consider a move.” One producer relaxed immediately. Until then, she’d felt as if a decision had already been made on her behalf.

When one person hesitates. The third producer said, “I don’t want the whole group to wait on me.” Morgan replied, “You can take the time you need. What would help you make your own decision?” They agreed to a product walkthrough. Morgan told Ellis that the branch would plan around confirmed starts and keep the remaining decision open. A leader can respect the team’s connection and still learn what each individual needs from the platform.

Use the same core questions with each person and leave room for their own priorities. “Where do you spend time doing work you expected someone else to own?” “What would be harder for you to replace than people assume?” “Who needs to be involved before you can decide?”

Summarize the answer back and ask permission before sharing individual concerns with a larger group. Morgan’s recruiting notes separated shared team requirements from personal decisions. If you’re the branch leader, use the same principle when people ask you to speak for your own team. You can describe how you work together and still give each person room to choose.

Separate the interview notes. Make a shared list for workflow dependencies and a restricted list for individual concerns. A shared dependency might be coverage for scenarios every producer uses. An individual concern might be schedule flexibility. In the group discussion, say which coverage the branch can demonstrate and invite each person to confirm whether it works for them. Before a recruiter sends a group summary, have them read it from the perspective of the quietest person on the team.

Show the work on a realistic file

Morgan could have booked a long tour of the platform. Instead, Ellis brought a fictional purchase scenario modeled on a difficult week. The scenario involved a tight closing and a late change to the documentation needed from the buyer. Priya showed where the file would go, which messages the producer could expect, and what would happen if the normal owner was out.

Then Morgan asked Ellis to describe how the group operated today. They found a place where Ellis’s team was already faster. Morgan wrote it down as a practice worth keeping during the move. That moment mattered because the conversation finally felt like two experienced teams comparing notes.

Run the messy version. Priya started the demonstration with everything going according to plan. Ellis asked for the version in which the closing time changes and the normal owner is out. Priya named the backup but discovered the escalation channel still listed the old contact. They corrected the job aid and asked the backup to run the same scenario. Ellis now had evidence of how the branch responded to a small gap, as well as the revised process itself.

Choose a scenario that represents the work the group actually does. Make it specific enough to expose a decision. Who owns the next action? What triggers a handoff? What does the producer see? Which promises need somebody else’s approval?

End the demonstration by naming the gap you discovered and who will answer it. Morgan’s follow-up was short. “You saw how Priya handles a deadline. We still owe you a clear answer on backup support. Priya and I will confirm that before we revisit the start date.” Specific follow-through builds trust even when the answer takes longer than expected.

Turn a demo into evidence. Ask the viewer what they saw, what they inferred, and what they still need verified. Capture all three. If a production leader says “you have great communication,” bring it back to what happened. “You received an owner and an update time when the first contact was out.” Then ask when that promise applies. Run one ordinary scenario and one exception. An honest demonstration can end with a limitation that both sides are willing to plan around.

Pick the order of arrival

The group wanted to start together. Morgan understood why. People who depend on one another usually want to know where the others will land. Priya still had a hard limit on training time. Morgan offered a common planning date, individual decisions, and a phased start with a clear reason for the order.

The first two producers would start when access, support ownership, and backup coverage had been checked. The third would follow once the new processor had demonstrated readiness and the first cohort had passed a service check. Ellis asked what would happen if the milestone slipped. Morgan added that answer to the plan.

Tell the group what a delay would look like. “If the backup isn’t trained by Friday, we’ll move the second start window and give you a revised date on Monday,” Morgan said. Ellis asked if the first two people would then be asked to cover the third person’s expected workload. Morgan said the branch would plan to the lower capacity until the milestone was met. The team could now decide with a clear picture of what would happen when the timeline slipped.

Make the sequence visible. Put access, training, scenario support, and first-file supervision on one page. Give each item an owner and a date. Mark which condition can delay a start and what you will tell people if it does.

A pre-start call is a good place to rehearse the first week. “Your first purchase comes in at ten Monday. Walk me through where you go for help.” If the answer relies on three people interpreting the same handoff differently, fix the handoff. Use approved transition channels and fictional file details for rehearsal.

Give the pause an owner. Put the milestone in a shared plan with one named decision maker. Record the evidence they’ll use, who communicates the decision, and who can cover a question if that person is out. Rehearse a delayed start notice in two sentences. Acknowledge the change and say when the next reliable answer will come. Ask the first cohort which commitments they made based on the original date so the team can address real impacts before they multiply.

Watch what changes after people arrive

Ten days into the fictional rollout, the first cohort had submitted fewer files than forecast. Morgan’s manager wanted a production update. Priya noticed the same two people were asking for help with product scenarios and weren’t sure how to reach the specialist. The issue appeared before it showed up in monthly funded volume.

Morgan added a short coverage review to the weekly meeting. Who had an open scenario, how long had it been waiting, and what could be handled without involving another person? Within a week the producers could get answers directly. The next coaching conversation could focus on business development instead of platform navigation.

Read the first numbers carefully. The cohort had five active files, two unresolved questions and one incorrect handoff. Morgan could have called that poor adoption. The notes showed the same queue was unfamiliar to both people and the instructions named two possible owners. Priya clarified the path, watched the next submission, and checked the handoff again a week later. The manager could then see whether the fix changed behavior without turning the coaching meeting into a lecture about activity.

Pick a few indicators you can act on. Morgan tracked first-file handoff, age of unresolved support requests, number of files entering the constraint stage, and confidence in the next two weeks of the pipeline. Funded volume remained a measure of the outcome. The earlier indicators gave the team a chance to change it.

Ask each producer, “What are you doing yourself that you expected us to handle?” That question catches hidden work. Ask existing team members what has become harder since the new people arrived. If growth is working for the newcomers and exhausting the people already there, you need to know before you commit to the next cohort.

Run a ten-day review. Compare your expectation with what the producers and existing team saw. Ask for one process that ran smoothly and one that depended on a person stepping in outside their role. Put both on the review agenda. If the rescue depends on someone staying late every night, include that in the support model. Close the meeting by deciding what will be documented, what needs training, and what should change before the next group arrives.

Branch readiness worksheet

Write down the historical production total and how you removed duplicate credit. Record your low, base, and high assumptions for eligible volume and funded units. In the next column, show the contribution assumptions, fixed support cost, and the margin for a slower ramp.

Then write one sentence for each of these: the operational constraint, the support owner, the backup, the condition that starts the next cohort, and the person who can call a pause. If any one sentence is vague, put that item on the next meeting agenda. A short list with real owners is easy to use on a busy Monday.

Work the example. Put 60, 48 and 36 funded loans in the base, lower and stress columns. Multiply each by the contribution that matches its scenario, then subtract added support. Add a separate row for a three-month ramp. Under the table, name the person who can verify each assumption. If a contribution figure is missing, keep it open. Showing which numbers still need an answer helps a branch manager get a better meeting with finance.

Close the readiness meeting with a decision everyone can repeat. “We’re ready for the first two once support access and backup coverage are confirmed. We’ll check the first ten days of service before committing to the third start.” Give the next decision a date.

After that meeting, consider what you would tell Ellis in private. A good answer should make sense to the producer and to the people who will serve the files. If you find yourself giving them two different explanations, you have one more conversation to have before promising a date.

Use a one-minute close. Say how many starts you can support, what you expect them to need, and how you will know when there is room for the next. Ask each operating owner to repeat their part. If the backup owner cannot explain what they do on a Thursday when the queue fills up, set another rehearsal. Morgan’s final condition was simple: every accepted start had a service owner, a backup and a review date that the producer could name.

Check the staffing decision again after two weeks. If the first cohort needs twice the coaching time you expected, ask which work can move and which service commitments must stay covered. Put that answer beside the financial model before you book the next start.

Your action sheet

1. Historical production and shared credit
2. Low, base and high funded-unit cases
3. Capacity constraint and coverage owner
4. First cohort conditions
5. Next decision date and pause criteria

Download the free PDF for the printable page.