Sample-data proof · Estimate changes · Human-reviewed AI

The scope-change loop keeps revised work from becoming a pricing mess.

A common home-service leak happens after the first estimate: the customer adds a room, removes a line item, or changes timing. This proof page shows a safe workflow where AI prepares the revision packet, but the owner approves scope, price, and customer wording before anything is sent.

Buy the Starter Kit — $29 See estimate handoff

The leak

Estimate changes often scatter across texts, photos, notes, and memory.

Scope drift

Customer asks for more

A quick text adds trim, ceiling, haul-away, or extra rooms after the original quote was drafted.

Pricing risk

Old assumptions stay hidden

Labor, material, access, or schedule assumptions from the first estimate do not automatically update.

Approval gap

Reply goes out too fast

The business answers before scope, price, timeline, and customer wording are checked together.

Sample revision queue

AI prepares the change packet. The owner approves the revised promise.

Customer change request“Can we add the stair rail and upstairs hallway?”
AI draftAttach to original estimate, identify added surfaces, flag missing measurements/photos.
Human approvalRequest missing info, schedule check, or approve revised scope.
Assumption reviewOriginal quote used one-coat wall refresh and no trim work
AI draftList assumptions that change labor, material, access, timeline, or margin.
Human approvalEdit quantities, pricing, timeline, and risk notes before customer response.
Customer-facing replyRevision needs clear next step
AI draftShort reply explaining what changed, what is still needed, and when the revised number will be ready.
Human approvalSend, personalize, call first, or mark the change out of scope.

The safety rule

AI can compare the new request against the old estimate, flag changed assumptions, and draft the reply. It cannot change price, promise schedule, or send revised terms without human approval.

What the Starter Kit teaches

Choose one revision trigger, define what the AI must check, create the owner approval card, and log each correction so the next scope-change packet gets sharper without removing the human decision.