“Closed every August lead in 7 days.”

The owner ran every lead through his phone. A homeowner would text a yard photo at 9pm, he would reply at 11pm, and a quote would land in their inbox three days later — typed into Word, exported to PDF, attached to a Gmail. Some leads waited a week for a quote because he was on a job site running install.

The shop closed deals — they just closed them slowly. The eighteen-day average from first touch to signed quote meant they leaked roughly one in four leads to a faster competitor. In peak season the owner was working 70-hour weeks just to keep the inbox empty.

We replaced the phone-as-CRM with Gladius CRM, wired the AI receptionist to the same number so inbound texts got an immediate operator-voice reply, and ported the quoting flow into the DMS-style desking surface — so a quote could be built, branded, and sent in under three minutes from any device.

We did not rebuild their workflow. We just moved it into one tool. The owner kept his voice — the AI receptionist drafts in his cadence, surfaces a one-line summary, and asks before sending anything that is not a routine reply. Every quote goes out from his sender address with his signature.

By the second week, response time on inbound leads dropped from a median of 4 hours to under 8 minutes. By the end of August the close window was sitting at 7 days for every lead that came in that month — measured by the platform, not anecdotally.

The owner is now sleeping at night. He glances at the morning brief on his phone over coffee, replies to two or three high-touch threads himself, and lets the cadence handle the rest. The shop took on more work in August than any month in their twelve-year history without adding a single hire.