Lead intake automation for a home services company
AIpplied Labs connected a home services company's online lead sources to its CRM, Jobber. Leads from Facebook, Instagram, Yelp, Thumbtack, and the company website now file themselves in 5 to 55 seconds, each one tagged with its source. Before, staff typed every one in by hand. The build took four weeks at a fixed price.
| Business | Home services company, owner-run, with a small office team |
|---|---|
| System | Lead engine ($3,000 fixed) |
| Sources connected | Facebook and Instagram lead forms, Yelp leads and messages, Thumbtack, and every website form |
| Time from inquiry to CRM | 5 to 55 seconds per lead |
| Share of the year's clients from now-automated channels | About 1 in 4 |
| Build time | Four weeks from access |
| Payment | Half on signature, half on acceptance |
The problem
Leads arrived in five places, and someone in the office copied each one into the CRM by hand. About 1 in 9 client records had no lead source at all, so nobody could say which channel was paying for itself.
What AIpplied Labs built
- One path for every online lead. Each one lands in the CRM as a lead, with the customer's details and their own words in a note. The team qualifies it there.
- Repeat customers stay on one record. A lead whose phone or email matches an existing customer goes onto that record. An uncertain match is tagged Needs review with a link to the possible match, so a person decides. Nothing merges on its own.
- Alerts. A message for every new lead, an alert when one needs review, and an alert if a source goes quiet, so a broken connection shows up the same day.
- Every lead carries its source, so the "not set" count stops growing.
- Everything runs on the company's own accounts, including hosting and the connection tools.
What stayed manual, on purpose
Google Local Services leads stayed manual. Its connection passes a phone number and a recording but no name or address, and the team already enters those callers during the call. Automating it would have added empty records. AIpplied Labs wrote that into the scope and left it out.
The result
In the first days after launch, real leads from the website, Facebook, and Yelp filed themselves in 5 to 55 seconds each, and a repeat customer landed on their existing record. A field-name difference in one source was found and fixed the same afternoon.