23% lift in monthly lead capture

The client estimated they were losing 15 to 20% of inbound leads to voicemail after hours and at peak call times. Two CSRs handled every inbound call; in summer, hold times stretched past 2 minutes and abandonment spiked.

Quoting was entirely manual. An estimator drove to the site, measured, came back to the office, built the quote in Excel, and emailed it 2 to 3 days later. By then the homeowner had often signed with a faster competitor.

They had no systematic way to track which leads converted, which were lost to slow response, or what their true cost per acquisition was by channel.

Voice agent. Deployed into their existing phone system through Twilio and integrated with ServiceTitan dispatch. It handles after-hours calls, qualifies leads, books appointments directly on the dispatch board, and escalates emergencies to the on-call manager.

Quote generation. Connected to their pricing engine and DocuSign. For standard jobs (driveways, patios, sidewalks) the system generates a quote from intake data within 2 hours of the first call. Non-standard jobs still go to the estimator, with the data pre-filled.

1. Weeks 1 to 2: diagnosis. We audited the call flow, mapped the revenue leaks, and interviewed the CSRs and estimators.
2. Weeks 3 to 4: architecture. We designed the ServiceTitan and Twilio integration and selected the modules.
3. Weeks 5 to 8: voice agent deployment and testing in production.
4. Weeks 9 to 12: quote generation module deployment.
5. Month 4 onward: measurement and ongoing optimization.

The engagement started as a 90-day fractional AI leadership engagement for diagnosis and architecture, then moved to module deployments for the voice agent and quote generation. Ongoing support and optimization continue at a flat monthly rate.