1212 COMMAND CENTER
Configuration

Settings

Everything that decides what good looks like for JG Hause Construction. Change the economics and every judgement in the tool moves with them, because the target is derived, not typed in twice.

What this client sells

A roofing and siding lead is not a software lead. This is where that difference lives.

$
%
%
$
Target cost per sale
$550
$110 a lead at 20%
Gross profit per sale
$4,200
$14,000 at 30% margin
Break even cost per sale
$4,200
The target leaves $3,650 of margin per sale.

Safety rails

The tool cannot exceed these, whatever it recommends.

%
$
The cap is a percentage of what this account is actually spending, recalculated from the last 30 days every time you open the feed, so it cannot go stale the way a budget figure on file does. At 15% and the planned $49 a day that would be about $7 a day; the Recommendations screen shows the live figure. Today the tool has used $0.
  • Anything marked spend affecting asks before it goes out, every time.
  • Dry run rehearses the whole flow and sends nothing.
  • Nothing reads as applied until the platform confirms it.
  • Every applied change carries the exact steps to reverse it.

Channels and data

Adding a channel or a client is an edit to one file, not a rebuild.

ChannelAccountStateMonthly budgetNote
Meta Ads1171989478475558Live$1,500Added to the Windsor connector on 2026-08-21, so this client finally has spend to divide its sheet leads by.
Google Sheetlead source1NNJ3my0pUetnIRYW1sYfIpSBioM4Cg1-s3lUfcliCKsLiveยทOne row per Meta Instant Form submission, read straight from the sheet as CSV. Meta seeds every new form with a test submission and that row is excluded. Not the sheet in the older notes: that one was built for the Messenger campaign and the weekly AccuLynx export, and has been empty since Jul 15.
V1 keeps these edits on this screen so the arithmetic can be checked against real client economics. V1.1 writes them to the client record, and the recommendation rules read them from there.