Platform leads measure an action
A submitted instant form shows that someone completed the configured action. It does not confirm that the phone and email are usable, the request fits the service area, the work is offered, the timing is viable, or the person will answer.
Treat the platform count as the top of a measured path. The office outcome provides the next evidence. Do not collapse every later failure into “bad leads” or assume that inexpensive volume is automatically productive.
Check delivery before judging quality
Verify that each submission reaches the correct system, creates one record, retains campaign context, alerts the responsible person, and receives a tracked next action. Open several records end to end rather than trusting a green integration status.
Look for delay, missing fields, duplicate records, unmapped answers, notification failures, and ownership gaps. A lead that arrives correctly but sits unworked is an operations failure. A lead that never reaches the CRM is an integration failure.
Separate the response stages
Track acknowledgment, first real attempt, successful contact, qualification, appointment, estimate, and booked job. Report “never contacted” separately from “contacted and unqualified.” That distinction prevents slow follow-up from being misread as campaign quality.
Review the distribution of response times and the unresolved tail. Averages can hide leads with no attempt. Segment business hours and after hours because the correct staffing and automation response may differ.
Improve the form without over-filtering
Ask only questions that change routing or the immediate sales decision: requested service, location, timing, and a practical project-fit signal where necessary. Long forms can reduce volume without proving that the remaining contacts are better.
Compare form answers with real conversations. If a question does not help the office respond, remove it. If one answer predicts disqualification, make the rule visible and keep an exception path for ambiguous projects.
Read campaign quality through booked outcomes
Compare campaigns using valid inquiries, contact rate, qualification, appointments, estimates, and booked work. Cost per lead remains useful for delivery monitoring, but cost per qualified opportunity or booked job is closer to the owner’s decision when the data is dependable.
Account for capacity and sales process. A campaign may reach qualified homeowners while appointment availability is poor. Another may produce conversations that do not fit the current job mix. Diagnose the stage where the decline begins.
Run a controlled improvement
Choose one change—routing, response ownership, form wording, audience, offer, creative, or follow-up—and define what should improve downstream. Preserve the existing event and campaign history so the comparison remains interpretable.
Review message quality as well as speed. An instant reply that cannot answer the customer’s question is not a completed response. Automation should acknowledge, organize, and escalate while a person handles service fit, scheduling, scope, and judgment.
- Confirm every lead reaches one visible queue.
- Assign an owner and backup.
- Track real contact and qualification reasons.
- Compare booked outcomes with campaign and creative.
- State attribution and sample-size limits plainly.