AI audit · SaveAround
40 leads a week was the ceiling on how many groups they could reach
SaveAround wanted to scale from about 90 markets to hundreds without rebuilding the call center it had before COVID. The plan traced exactly where the manual work capped growth, and staged the fix into three build-ready phases.
1 / 5
The problem
Leads came from field reps, contractors, and a small call center, each finding their own groups by hand. When a contact wasn't listed publicly, someone dug through LinkedIn or Facebook for the band director or the coach, then copied results into a spreadsheet before hand-entering them into the CRM. A strong rep could build around 40 new leads a week that way, and that manual middle was the real ceiling on how many groups the business could reach in a season.
The team already trusted AI where it worked: a yearly merchant-verification job caught closed businesses and saved hundreds of hours. Nothing pointed that same capability at new-lead sourcing yet.
What we proposed
The plan staged the build around what SaveAround already had: a homegrown CRM with API access and two in-house developers, ActiveCampaign already running email, and a track record of trusting AI on the merchant side. Phase 1, targeted to be live before the September selling season, replaces the manual sourcing outright: a plain-English request finds the groups and organizations, pulls the contact for that role from public listings or LinkedIn and Facebook when it isn't public, checks it against the roughly 60,000 leads already in the CRM to avoid rework, and writes the clean new leads straight in.
Phase 2 turns the leads Phase 1 produces into conversations: a personalized first email referencing the group, role, and area, a spaced three-to-four email rhythm, then an AI voice follow-up call for the ones who go quiet, since phone is what tends to get a response once email doesn't. Phase 3 closes the loop by re-verifying the existing 60,000 contacts once a year, the same way the merchant list already gets checked, so renewal season starts with a clean, warm list instead of a stale one.
The result
The manual middle was the ceiling. The plan removes it in three phases.
40
new leads a week, the hand-built ceiling per rep
60,000
existing leads the new engine dedupes against
3
phases, staged to be live before the September season
What the client said
“★★★★★”