The registry

The Ledger.

Everything Capstone publishes, filed in one place: numbered Field Reports and Field Notes from seven years of operating marketing inside an eight-figure home services company.

We publish what this work costs and what it returns, because the numbers are the argument.

From the operating record

Field Notes

Real ledgers, real mistakes, no vendor numbers.

FN-01

The Best Channel I Ever Ran Had No Dashboard

Clipboards, a van, and a spreadsheet: $1.99M at 15.5% fully loaded. The complete 2021 ledger of a door-to-door program.

FN-02

The Market Did $2M That Year. We Still Killed the TV Campaign.

Traced by source, the channel cost more than the jobs it produced were worth, and the kill line took it.

FN-03

How We Cut the Aggregator in Half (And What We Built First)

The three-year sequence, 2019 to 2021, that made HomeAdvisor redundant instead of removed.

FN-04

We Kept Knocking Through COVID

Three weeks of Zoom homework, then back to the doors: $2.2M at 12% COM through a pandemic.

FN-05

A Storm Saved Our Year. Two Elections Wrecked Two Others.

Same company, same systems, three year-end numbers swung by events no marketing plan touched.

FN-06

What a Lead Costs in 2026, by Trade

$39 electrical, $51 HVAC, $57 plumbing on LSA; $183 in non-branded search. The data, and the calculation that turns it into a decision.

FN-07

Your Best Channel Is Lying to You

Strip your company name out of the campaign and the true acquisition cost often doubles. The twenty-minute check.

FN-08

Plumbing Isn’t One Market. It’s Four.

Emergency, scheduled repair, replacement, big-ticket: four funnels, four timelines, four different winning channels.

FN-09

The Three Days Your Truck Is the Best Ad in the Neighborhood

Every active job site is a temporary broadcast tower of third-party proof. The four-step harvest most contractors let expire.

FN-10

Our Lead Count Fell by Half. Conversion Nearly Doubled.

Six years of data on what happens when the dashboard’s favorite number stops being the goal.

FN-11

Facebook “Didn’t Work” for Four Years. Then We Built a System.

Every owner who says the channel failed is right about their experience and wrong about the channel. What changed when the system did.

FN-12

Our Cost Per Lead Quadrupled. The Business Got Healthier.

$85 to $371 over five years, and why that chart is a sign of health, not a firing offense.

FN-13

In Detroit, Boots Beat Broadcast by 60 Points. They Still Weren’t Good Enough.

The other half of the Detroit expansion: the canvass ledger, side by side with the broadcast campaign that got killed.

FN-14

We Put Our Prices on the Website. It Closed $342K.

The three objections every contractor raises about publishing prices, and the ledger from doing it anyway.

FN-15

It Launched at 137% COM. Fourteen Demos Is Not a Verdict.

Our own kill bands say cut anything over 50%. Why this launch got sample size instead of the axe.

FN-16

A Demo Is Not a Demo

A 55.1% close rate that is real and also a fiction. What the single number hides about the demos underneath it.

FN-17

We Stress-Tested Losing Our #1 Channel. Half of It Was On Purpose.

In 2021, 784 demos rode one channel. What we did about the concentration before it did something to us.

FN-18

We Upgraded to Higher-Quality Leads. I’m Still Not Sure It Was Right.

The 2024 rebuild of the door-to-door channel, and the volume-for-quality tradeoff that still is not settled.

FN-19

Five Call Center Managers in Five Years. The Seat Finally Taught Us How to Fill It.

Every lead dollar funnels through one seat. What five hires taught us about who belongs in it.

FN-20

One Vendor Was 95% Brand. The Other Was 100% Discount. Both Were Wrong.

Two broadcast vendors with two opposite religions, and where the working mix actually sits.

FN-21

The Cheapest Channel on Our Books Took Five Years to Build

500 Google reviews became 1,620 while organic revenue grew from $676K to $1.21M a year, at near-zero media cost.

FN-22

A Demo Costs $992 Now. It Is the Best Money on the Sheet.

Cost per demo nearly tripled from $337. Close rate and average ticket rose faster, and the appointment got more valuable.

FN-23

We Cut Aggregators for Five Years. This Year We Raised Them on Purpose.

Share of budget fell 22.6% to 9.3%, then went back to 14.5% by choice. The difference between a dependency and a dial.

FN-24

We Set 9,548 Appointments. Reps Got 3,993 of Them.

The funnel stage with no media budget attached to it, and why our best revenue year had the worst intake discipline.

FN-25

A Channel Only One Person Can Run Is Not a Channel

A division that already worked collapsed months after one person changed seats. What it took to rebuild it as a system.

FN-26

Two Halves of Our Calendar Ran on Different Clocks. Nobody Noticed.

Canvass booked two to three months out while the call center booked next-day. Which metrics get orphaned, and why.

FN-27

The Best Hire I Ever Made Interviewed Badly

Interviews test the traits that make someone good at interviews. How to test the trait the seat actually requires.

FN-28

We Were Certain Someone Was Stealing. It Was a Missed Step in a Process.

Distance from the work makes personnel theories feel obvious. The four questions that come before any of them.

Every entry in this registry traces its figures to a cited source or to our own operating records. Errors flagged are corrected and logged. The operating record behind these entries is filed in The Record.

Want your own ledger read?

One page of findings on your marketing, held to the same standard as everything filed here. Free, no meeting required.

Or call (419) 575-9023