Operating advisory for lending executives ◆
Most companies don't have a strategy problem. They have an operating problem.
Twenty-five years in mortgage and lending — wholesale, correspondent, TPO and digital. I've built channels from nothing, scaled monthly production from $20M to $250M, and pulled $7,800 of cost out of a loan without touching capacity. Now I do that for one executive team at a time, across every function rather than the one that hurts.
Every engagement starts the same way: map the process flows as they actually run, across all five systems and through the seams between them. Once a flow is on paper, two things become obvious — what to fix before touching any tool, and where an AI tool would earn its cost rather than automate a broken step faster.
Track record ◆
Twenty-five years, ten organizations, and a lot of broken pipelines.
Wholesale, correspondent, TPO, retail and digital. Startups built from nothing and $6B platforms rebuilt while they ran.
Led the transformation of a non-QM shop into an Agency lender — 1,150% growth in monthly production, a 190-person support organisation built to carry it, and $120M+ in net revenue generated.
Origination cost taken out at a national bank without cutting the capacity to fund. Cost per loan is an operations number, not a pricing one.
A 131% efficiency improvement on volume that was already coming in the door. The lending version of the seam between sales and fulfilment.
Across Freedom Mortgage and MetLife Bank — 120-person operations consistently clearing $500M in monthly production across brokers, bankers and credit unions.
Case study 01 · Quontic Bank ◆
Everyone was looking at underwriting. The constraint was the front door.
Pull-through sat at 32% — two out of three files never closed. The instinct in the building was to go get more brokers. That would have made the problem bigger and more expensive.
Intake and stacking were being done by hand. Every file in the building queued behind that one step before anyone underwrote anything — so conditions went out late, and late conditions are what actually kill a loan.
What followed was predictable. Brokers chased status by phone. Resubmissions landed at the back of the same manual queue. Files aged, rates moved, and deals repriced or died. Underwriting took the blame for a delay that had already happened upstream of it.
Pull-through is the cleanest seam metric in lending — the only number measuring what sales promised against what fulfilment delivered. It sat between two org charts, which is precisely why nobody owned it. Meanwhile 42 monthly active brokers were carrying the plan, and the good ones were sending their next file wherever they got a faster answer.
The front door first. Intake, stacking and document classification automated, so files stopped queuing before they started. Conditions moved to submission rather than day five, and every stage got a published turn time the sales floor could quote with confidence.
Broker quality made measurable. Submission standards and scorecarding, because one broker who submits clean is worth three who don't. The network then scaled from 42 to 200+ monthly active clients — growth second, deliberately.
The sales organisation rebuilt. 97 sales executives recruited and mentored against real training and performance systems rather than activity metrics.
Systems consolidated. HubSpot managed end to end with dashboards leadership actually ran the business from. The $7,800 per loan came from two places: taking out the vendor and technology sprawl, and killing the fallout that was carrying full origination cost on files that never funded. All of it inside OCC compliance across every channel.
A 131% efficiency improvement on volume the bank was already paying to acquire. The gain started at the front door and compounded through every stage behind it.
Two sources: vendor and technology consolidation, plus the fallout that was carrying full origination cost on files that never funded.
Growth came second, deliberately. A wider network only compounds when the operation behind it can convert what shows up.
Built on training and performance systems rather than headcount alone — and inside full OCC regulatory compliance across every channel.
$20M to $250M a month. The operation had to exist before the volume did.
A non-QM shop with an Agency ambition. Scaling production more than twelve times over in under three years is not a sales problem — it's a build-ahead problem, and almost everyone builds behind.
Growth at that rate breaks anything not designed for it. Fund faster than fulfilment can absorb and pull-through collapses, brokers leave, and the growth destroys the margin it was supposed to create.
The sequencing mattered more than the ambition: 190 support staff and the fulfilment infrastructure went in alongside 120 wholesale AEs, not after them — and all of it inside CDFI compliance and an Agency approval process that doesn't bend for a growth plan.
1,150% growth in monthly production and $120M+ in net revenue, generated through fulfilment operations and warehouse partnerships rather than volume alone.
Alongside it: a Federal Home Loan Bank relationship established, the digital portal rebuilt, and the brand repositioned for the lender the company was becoming instead of the one it had been.
$20M to $250M, non-QM to Agency lender, in under three years.
Fulfilment capacity stood up alongside 120 wholesale AEs, not trailing them.
Through fulfilment operations and warehouse partnerships, inside CDFI compliance.
Achieved without the margin collapse that usually accompanies growth at this rate.
How it works ◆
Diagnose first. Nobody needs another deck of recommendations they can't staff.
The teardown
2–3 weeksAll five systems, assessed together. Interviews with your leadership team and the people two levels below them, plus your actual numbers — not the board version. You get a written findings memo that names the constraints in priority order and what each one is costing you.
The operating plan
2 weeksA sequenced plan with owners, dates and a defined measure of done. Sequenced deliberately: the things that fund the rest go first, and nothing lands on a team that doesn't have the capacity to absorb it.
Execution support
3–9 months, optionalI stay on as an operating partner to your team — weekly cadence, unblocking, and hard conversations when the plan meets reality. Or you run it yourself and I check in monthly. Both work; the choice is yours.
On AI ◆
Ask executives what worries them about AI, and almost nobody says the technology.
They say governance. Accountability. Whether their people are about to walk. Whether any of it will show up in the numbers. Those aren't engineering problems — they're operating problems, and they're the ones I'm hired for.
Risk, oversight & governance
Not a technology gap. Nobody has decided who approves an AI-assisted decision, what gets reviewed before it reaches a customer, what gets logged, and whose name is on it when a regulator or a client asks. The tools arrived before the controls did.
What I do about itPut the operating controls in before the tools go further: decision rights, review thresholds sorted by risk tier, an audit trail that already fits how you work, and a named owner for every use case. Governance that fits on one page, because the twelve-page version doesn't get followed.
Employee fear of replacement
An unanswered question. Leadership stays quiet while it decides, and people fill the silence with the worst version. Your strongest performers — the ones with options — start taking calls before you've made any decision at all.
What I do about itGet leadership to an actual position and say it out loud: which roles change, which don't, and what happens to the capacity that gets freed up. Then redesign the roles around the work that's left, so the answer is visible in the org chart and not just in a town hall.
Where to deploy & adoption
A prioritisation failure dressed up as a technical one. Most teams point AI at the most visible process rather than the most expensive one, pick something with messy data underneath it, and then can't understand why nothing moved.
What I do about itRank the use cases by volume and cost-to-serve against how ready the data actually is. You get a short sequenced list — and, just as usefully, a written list of the ones to refuse, so they stop coming back every quarter.
Unknown or unmeasurable ROI
No baseline. If nobody measured cycle time, error rate or cost per unit before the pilot started, there's no way to prove anything afterwards — so the pilot quietly dies at budget season and takes the team's appetite with it.
What I do about itBaseline first, always. Every use case gets a before-number, a target, a decision date and one person accountable for making the call to scale it or kill it. Killing it on schedule counts as a result.
Three of the four are operating problems wearing a technology costume.
That's the whole argument for reviewing the business as one system. You can't govern what you haven't mapped, you can't measure what you never baselined, and you can't calm a workforce with a position leadership hasn't taken yet.
Source: industry poll of business leaders, 2026. Replace with your own figures and attribution before publishing.
Where it earns its keep
- High-volume, low-judgment work that's currently absorbing headcount you'd rather deploy elsewhere
- Reconciliation, triage, classification and first-draft work with a clear quality check on the other side
- Anywhere your people spend hours assembling information that already exists somewhere in the business
- Roles you were about to hire for — before you commit to the fixed cost
Where it's theatre
- Anything sitting on data three teams already report three different ways
- Processes nobody has mapped — you'll automate the broken version faster
- Work where being wrong 5% of the time is a regulatory or customer problem, with no review step budgeted
- Pilots with no owner, no baseline and no decision date attached
Who you'd be working with ◆
James Clayton Hooper
Twenty-five years in lending across ten organisations — wholesale, correspondent, TPO, retail and digital — almost all of it as the executive accountable for the number rather than the one advising from outside it.
I've built a wholesale lending entity from nothing: 120 account executives, 190 support staff, and the compliance and operations infrastructure underneath them. I've taken a non-QM shop to Agency lender and scaled it from $20M to $250M in monthly production. I've run $6B+ in annual volume at Freedom Mortgage and MetLife Bank. And I've done the unglamorous version — pulling $7,800 of cost out of every loan at a bank that had the volume and couldn't convert it.
What I kept finding is that the problem is almost never where the org says it is. Sales blames operations, operations blames sales, and the constraint is sitting in the handoff between them where nobody owns it. That's why the teardown looks at all five systems at once, and why the first number I ask for is usually pull-through.
I work with a small number of companies at a time. If we're not a fit, I'll tell you on the first call and point you somewhere better.
- Best fit
- Mortgage lenders, banks and credit unions running wholesale, correspondent, TPO or retail channels. Also mortgage technology platforms selling into them.
- Been there
- Freedom Mortgage · MetLife Bank · Quontic Bank · Change Home Loans · TMS · NMSI · Mandalay Mortgage · Equifirst · US Bank.
- Systems
- Salesforce (Admin & Design certified), HubSpot, LOS integration, NMLS and CoreLogic data. Built a custom Salesforce CRM integrating LOS and NMLS data from scratch.
- Recognition
- NMP Top 40 Under 40 (2015, 2016) · NMP Most Connected Mortgage Professionals (2017) · Consecutive Regional VP of the Year, US Bank.
- Engagement
- Fixed fee for the teardown and operating plan. Monthly retainer if you want execution support.
- Capacity
- Three engagements at a time.
Revenue & channel
- Wholesale, correspondent, TPO and retail channel build
- Broker network growth and AE recruiting at scale
- Sales performance systems, training and comp design
- Go-to-market strategy and enterprise partnerships
- P&L ownership
Operations & cost
- End-to-end process flow mapping across sales and fulfilment
- Cost per loan and cost-to-serve reduction
- Pull-through, fallout and cycle time
- Fulfilment operations and capacity planning
- Org design, span of control and decision rights
- Regulatory compliance — OCC, CDFI, Agency
Systems, data & AI
- Salesforce builds — Admin & Design certified
- HubSpot managed end to end, with performance dashboards
- LOS, NMLS and CoreLogic data integration
- AI use-case assessment across every function — where it takes out cost, and where it's theatre
- Governance and controls established before deployment, not after
Book a consultation ◆
Start with a 45-minute call.
No charge, no deck. You describe what's not working; I'll tell you what I'd look at first and whether this is worth either of our time.
- I'll come having read whatever you send in advance
- You'll leave with at least two things you can act on regardless of what happens next
- If a teardown makes sense, I'll scope it in writing within three days