The record

Ideas are cheap.
The numbers
have to move.

Four operating turnarounds across four industries. Same pattern every time: find where the system comes apart, redesign how work moves, then measure whether it actually worked.

Companies are described rather than named — several engagements carry open-ended confidentiality. Names available in conversation.

Case 01 · Healthcare

A national health insurance market

Recruiting, licensing, onboarding, compliance, lead economics, coaching and execution all had to move together — or the season was lost.

The problem

Enrollment targets required producing licensed, trained, compliant agents faster than the onboarding pipeline could physically deliver them. Every function was optimizing its own step while the calendar ran out.

The change

Rebuilt the operating cadence end to end — one sequence from recruit to producing agent, with owners at each handoff, lead economics tied to capacity, and coaching triggered by live performance instead of monthly review.

The result

The market finished first nationally, above enrollment and revenue goals, with onboarding time cut by 40%.

#1Nationally
148%Of enrollment goal
+32%Above revenue target
10→6 wksOnboarding

What transferred: when a deadline is fixed, speed comes from removing handoff latency — not from working the individual steps harder.

Case 02 · Medical devices

A medical-device business with a CRM nobody trusted

Tens of thousands of records, and almost no operating intelligence. The pipeline was a guess with a spreadsheet attached.

The problem

78,000+ records nobody believed. Reps kept private spreadsheets, forecasts were negotiated rather than measured, and leadership couldn't tell which activity actually produced revenue.

The change

Rebuilt the information architecture: lifecycle stages that matched how deals really moved, attribution that survived scrutiny, workflows that captured data as a byproduct of selling, and reporting leaders could act on the same day.

The result

Revenue doubled inside twelve months, with enterprise agreements built on a pipeline the team could finally defend.

$4M→$8MRevenue, 12 months
$7.5M+Enterprise agreements
78,000+Records, storage → system

What transferred: a CRM isn't a database, it's an operating system. If people work around it, the data is fiction and so is the forecast.

Case 03 · Home services

A growing service company losing margin quietly

Revenue was moving. The operating model underneath it was leaking money through handoffs and untracked spend.

The problem

Top-line growth masked bottom-line erosion. Operations and finance were disconnected, spend was approved locally with no aggregate view, and nobody owned the total cost of delivering a sale.

The change

Centralized operations, connected finance to sales activity, automated the workflows that were being done by hand, and put a real owner on cost per completed job.

The result

Inside ninety days: growth on both lines, and $300,000 of recoverable spend surfaced that nobody had been able to see.

+10%Top line
+20%Bottom line
$300,000Identified savings, <90 days

What transferred: margin rarely disappears in one place. It leaks in small amounts across seams nobody owns.

Case 04 · Distributed sales

A sales organization that had to triple without breaking

Scaling wasn't a hiring problem. Territories, leadership, training, accountability and launch operations all had to scale together.

The problem

Aggressive market expansion with a structure built for a third of the size. Hiring faster would have produced more untrained people failing in more places.

The change

Built the scaffolding first — territory design, a leadership layer that could absorb new reps, a repeatable onboarding playbook, and launch operations that made market number ten look like market number two.

The result

40 to 120+ people across 10 markets in under three months, without the quality collapse that usually accompanies that curve.

40→120+People
10 marketsOpened
<3 monthsTimeline

What transferred: structure before speed. Hiring into a broken model just distributes the breakage.

The one that taught me the most

The automation that worked, then didn't.

I built an AI workflow that cut administrative work after customer calls dramatically — three hours a day back to agents. It worked. So we pushed further and let it write and send the follow-up communication too.

Conversion dropped. Not because the writing was bad — because at scale, efficiency became distance. We put human judgment back into the final message and performance returned.

The machine wasn't the problem. The allocation was.

This is where The Human Gradient came from

Different industry. Same job: build the operating system behind the growth.

Start the conversation See the operator card