Case study: $710K in revenue, $0 ad spend โ€” a regional home builder | PrimexAI
CASE 04 ยท B2C ยท Real estate

Home builder: $710K in revenue, $0 ad spend

A database of 5,200 dead inquiries had sat untouched for 2 years. A WhatsApp AI agent + AI voice agent woke it up in 6 weeks โ€” 440+ leads a month, not a dollar on ads.

$710K
in revenue
440+
leads/mo (was 80โ€“120)
8
homes sold
$0
on ads

The situation

Regional home builder ยท single-family & townhomes ยท 3 communities for sale

For 2 years the builder had been collecting leads โ€” people who'd inquired about buying a home but never bought. 5,200 contacts. Calling them by hand was hopeless, and the ad budget had been cut. The database just sat there going stale.

Before the rollout
  • 5,200 warm leads up to 2 years old โ€” not one got a systematic follow-up
  • Ad budget slashed: paid channels delivered just 80โ€“120 leads/mo at a high cost per lead
  • Working the list by hand was impossible โ€” no time, no headcount
  • No idea who in the database was still active and ready to buy now
  • 3 separate communities โ€” each with its own buyer, so one blast to everyone wouldn't work
Before / After
BEFORE ยท Leads per month
80โ€“120 with paid ads running

AFTER ยท Leads per month
440+ from a dead database

BEFORE ยท Ad budget
Thousands of dollars a month

AFTER ยท Cost to acquire
$0 in ads

BEFORE ยท Database status
Dead โ€” 2 years with no contact

AFTER ยท Revenue from the list
$710K in 2 months

What we did

A two-step funnel: a soft touch over WhatsApp first (higher open rates, less resistance), then an AI voice agent to call back everyone who replied.

1

Segment the database by community and buying stage

We split the 5,200 contacts into 3 segments by community, then by budget (from intake forms and conversations up to 2 years old) and buying stage: "just browsed," "toured a model home," "requested paperwork." Each segment got its own message and script.

2

WhatsApp AI โ€” first touch over messaging

The first contact was a WhatsApp message, not a call. A short, personalized note: "Hi [name] โ€” you looked at homes in [community]. We have new listings available; would you like to take a look?" WhatsApp open rates run 3โ€“4ร— higher than SMS.

WhatsApp AI
3

AI voice agent โ€” qualify everyone who replied

Anyone who answered "yes" or "interested" on WhatsApp got an automated AI voice call within 15 minutes. The agent qualified them on budget, timeline, home specs, and readiness to tour. Hot leads went to the top of the queue.

AI voice agent
4

Full-context handoff to the sales rep

The rep got a complete lead record in the CRM โ€” name, community, budget, timeline, objections, preferred tour time โ€” before the first call. The conversation didn't open with "tell me about yourself," but with "you looked at [community] โ€” want to come by Saturday at noon?" Conversion to a booked tour jumped several times over.

After 6 weeks

$0 spent on ads
$710K
Total revenue in the first 2 months โ€” from a database that had been written off as dead for 2 years and brought in nothing
440+
Qualified leads a month โ€” versus 80โ€“120 with paid ads running. A 3.5โ€“5ร— increase
8
Homes sold in the first 2 months โ€” from a list of people who supposedly "would never buy"
$0
Spent on ads. Every lead came from the builder's own database. Cost per lead a fraction of any paid channel

More case studies

Sitting on a dead database?

Get a free diagnostic: we'll size up your list, forecast the leads it can produce, and show how much revenue you can pull out of it with zero ad spend.

Get a free diagnostic โ†’