Operating Record

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

In 2021, door-to-door was the backbone of the demo calendar — 784 demos, the single biggest source in the book — and our top three sources together produced 51.4% of everything. A bad season for the doors would have been a bad year for the company, full stop. Most contractors live with a version of that exposure their whole careers and never measure it.

Five years later the doors contribute 91 demos year-to-date — a fraction of their old volume — and the company is converting at all-time highs. We effectively ran the disaster drill every owner fears. The business passed.

The honest part: this wasn’t all strategy

The clean version of this story is “visionary leadership deliberately diversified.” The real version: the door channel’s decline came from a deliberate restructuring toward fewer, higher-quality leads — a trade whose conversion side has clearly paid and whose volume cost was steeper than planned, an honest accounting we’re publishing separately. So: half chosen, half learned. What was fully chosen — years before the decline arrived — was building the replacements while the backbone was still healthy.

Where the demos come from now

Four of the sources feeding the 2026 calendar did not exist as tracked lines in 2021: an annual-inspection membership program (123 demos YTD, closing at 76.5%), a website pricing calculator, a rebuilt social channel, and two newer aggregators run on a short leash. The mature channels filled in around them — Google organic, paid search, broadcast, referral. No single source is load-bearing anymore: the top three now carry 37.8% of demos, down from 51.4%, while the close rate sits at an all-time high of 59.8%. Less concentration, more conversion. Those were supposed to be a trade-off.

Concentration is the risk metric nobody tracks, because the months it matters are the months it’s too late to fix.

Why you build the second engine while the first still runs

Every replacement source on our 2026 calendar took years to mature: the membership program needed a customer base and a service rhythm before it could feed demos, the social channel failed for four years before it worked, the calculator produced three demos its entire first year. None of them could have been summoned in a crisis. If we had waited for the backbone to crack before starting them, the gap between the old engine dying and the new ones arriving would have been measured in layoffs. Instead it was measured in a concentration chart drifting calmly downward.

Run your own number tonight

One division: demos (or jobs) from your top three sources last year, over total. Above 60%, you’re one algorithm change, platform repricing, or key-person departure away from a crisis — and the time to start the next source was last year. Between 40 and 60, pick the one replacement engine that takes longest to mature and start it now, while you don’t need it. Under 40 with conversion holding: you’ve built what most contractors only have until the day they suddenly don’t.

We lost most of our #1 channel and posted record conversion in the same stretch. Not because losing it didn’t matter — because by the time it happened, it didn’t have to.