The five numbers that decide where the week goes. Click one for the math — and the work behind it.
Universe: all demand. Every one of the 612 inbound contacts in the base book — calls, forms, and LSA leads from every source, counted once at the door. (Installs, below, counts only the paid-campaign slice of this demand.)
Share of daily demand by hour. Amber = after hours (6p–7a), when the phones roll to voicemail.
Reason-coded losses for demand that never became a booked job. Click a reason to see the definition and the recovery math.
Demand we can still win back this week, oldest first. Worked top-down, these are worth real gross profit.
| Household | What happened | Est. GP | Waiting | Next step |
|---|
Recovery targets are re-contacted by a human within one business day. No blast texts.
Revenue minus parts, loaded labor, and callback cost, divided by paid hours. The only ranking the owner should coach to. Click a technician for the full build-up.
| Technician | Revenue | Net $/paid hr | Drive time | Callback rate | Jobs |
|---|
Every paid hour, split four ways. Drive and callbacks are the two leaks an owner can actually schedule away.
Darker = more callbacks. A column that lights up is a coaching conversation; a row that lights up is a process problem.
Universe: paid campaigns only — the 418 leads that began as ads, a subset of the Front Door's 612 contacts, never a second count of them. (Example: Paid Search shows 96 contacts at the Front Door, but only the 63 replacement-intent Replace leads enter this chain.) Platform ROAS stops at the click; this stops at the bank.
What one lead is actually worth after the job is done and paid for. Click a campaign to cross-filter.
Quoted work waiting on a yes. An estimate nobody touches for two weeks is usually already sold — by somebody else.
Two touches minimum: a call, then a text with the quote attached. Click any row for the follow-up script math.
| Household | System quoted | Quote | Age | Last touch |
|---|
Installed base aged 13+ years, weighted by repair history and plan status. Click a zone to cross-filter the whole cockpit.
Households most likely to need a replacement in the next 30 days — grouped by zone so one truck works one neighborhood.
| Household | Signal | System age | Est. value |
|---|
Open install slots, and the order they should be filled: plan members first, then past-due replacements, then repair-heavy households — new ad demand last.
Plan price minus the real cost to serve (visits, parts, drive, callback share). Click a cohort to cross-filter.
Cohorts are set by visits in the trailing 12 months: Light ≤2, Standard 3–4, Heavy 5+.
Agreements coming due, and how much of that revenue is on members who currently lose money.
Every member whose cost to serve beats their plan price. Each one gets a named fix: reprice, convert, or re-route. Nobody gets quietly cancelled.
| Member | Zone | Visits/yr | Cost to serve | Plan price | Net/yr | Fix |
|---|
Invoiced revenue, step by step: did it happen, and do we know where it came from?
A view that cannot reconcile to invoicing is labelled sample or dropped — never estimated into existence.
The short list where the data and the real world disagree. Cleared weekly, by a person, before the Monday numbers go out.
| Item | Amount | Owner |
|---|