I've spent the last few years running Meta & Google ads across real estate and local business niches, and the same mistake shows up almost every time I audit a brokerage's ad account: they're running ONE campaign for the whole brokerage instead of splitting spend by intent.

Here's the breakdown that actually moves the needle:

1. Separate buyer leads from seller leads.

A buyer lead campaign and a seller lead campaign should never share an ad set. Buyers respond to home search/browse creative. Sellers respond to "what's my home worth" valuation offers. Mixing them tanks your cost-per-lead on both.

2. Google Search for sellers, Meta for buyers.

Sellers are actively searching ("sell my house fast [city]") — that's high-intent Google Search territory. Buyers browse passively — that's Meta's strength with lookalike + interest targeting on active listings.

3. Route leads to a person, not a form.

If a lead sits in a CRM for 2+ hours before a human touches it, your close rate drops by more than half. The brokerages that win route leads via instant SMS/call to whichever agent is "on rotation."

4. Kill campaigns before day 14, not day 30.

Most agencies let underperforming campaigns run a full month before optimizing. By day 14 you have enough data — if cost-per-lead is 2x your target, cut it and reallocate, don't wait it out.

Ran this playbook across brokerages with 5-20 agents and it consistently cuts cost-per-lead by 30-40% within the first 60 days. Happy to answer questions if you're running your own ads or thinking about it.