Every real estate agent asking about “AI cold calling” is really asking one of two things: how to write a better script, or how to get more of those calls to turn into appointments. AI can help with both, but not in the order most people assume. The script is the easy part; any tool writes a passable expired or FSBO opener in ten seconds. What actually moves the needle is two things: knowing something real about the house and the owner before you dial, and actually picking up the phone.
What “AI Cold Calling” Actually Means for Agents Right Now
For most agents, AI cold calling breaks down into three separate tools doing three separate jobs: script and objection-handling generation, data enrichment on your call list, and AI-assisted dialing or voicemail drop. This sits inside the broader AI stack most agents are already building, and agents who feel like AI “isn’t working” for their calling usually only adopted the first piece of it.
A generated script sounds like every other generated script, because every agent using the same prompt gets close to the same output. Enrichment is where the differentiation actually happens, because it’s specific to the property and the person on the other end of the line, and that’s the part a generic prompt can’t fake.
Why Scripts Were Never the Real Problem
Expired and FSBO scripts have been public for two decades. Every brokerage has a swipe file. The reason cold calling still feels hard isn’t a wording problem: agents are dialing with the same three data points everyone else has (address, expired date, maybe a list price) and saying something a homeowner has already heard from the other four agents who called that morning.
AI doesn’t fix that by rewriting the opener. It fixes it by giving the agent something to say that the other four callers didn’t have.
The Move Most Agents Skip: Enriching the List Before You Dial
Founder’s Take: Most “AI for cold calling” content stops at ChatGPT writing you a script. That’s the easy 20%, and it’s basically solved. The part that actually changes outcomes is enriching the list before anyone picks up the phone: equity position, how many times the price dropped, who the prior agent was, whether the seller already tried FSBO before. That’s a data enrichment problem, and it gets skipped because “data enrichment” sounds like a SaaS term instead of a real estate term. It shouldn’t. It’s the single highest-leverage AI move available to an agent working expired and FSBO right now.
Before a single call gets made, a basic enrichment pass should answer:
- What’s the owner’s estimated equity position?
- How long was it listed, and how many price reductions happened?
- Who was the listing agent, and did the listing just expire or get withdrawn?
- Is this an absentee or out-of-state owner (which changes the whole approach)?
- Has this address come up on any other list (FSBO, expired, pre-foreclosure) before?
Real Estate-Specific Tools vs. General Data Enrichment Platforms
Agents usually stop at real estate-specific tools, which is fine as a starting point but leaves a layer of enrichment on the table that comes from outside the industry entirely.
| Tool | Type | What it pulls | Best for |
|---|---|---|---|
| REDX | Real estate-specific | Expired/FSBO lists with phone numbers, built-in Vortex dialer | Fast volume dialing on pre-built lead lists |
| PropStream | Real estate-specific | Ownership history, equity estimates, comps, absentee-owner flags | Deep property and equity research before calling |
| BatchLeads / BatchDialer | Real estate-specific | Skip tracing, list stacking across sources, built-in power dialer | Combining several lead sources into one enriched list |
| Clay | General data enrichment | Waterfalls dozens of data sources (public records, contact data, business ownership) into one enriched row per lead | Agents or teams who want custom enrichment logic beyond what RE-specific tools offer out of the box |
| Apollo.io | General B2B enrichment | Contact and firmographic data, mainly built for people and businesses | Investor-focused agents calling LLC owners or absentee corporate landlords |
The real estate-specific tools are built for volume. The general-purpose enrichment platforms (Clay in particular) exist because sales teams outside real estate solved the “know more before you dial” problem years ago, and almost none of that thinking has crossed over into residential cold calling yet. Piping a PropStream or REDX export through Clay to layer on ownership and contact data most agents never see is a genuinely underused move.
Before and After: How Enrichment Changes the Opening Line
Generic expired script (no enrichment):
“Hi, is this [Name]? This is [Agent], I’m a local agent and I noticed your listing expired. I’d love the chance to sell your home. Do you have a few minutes?”
Enrichment-informed opener (same lead, after a data pass):
“Hi [Name], this is [Agent]. I saw your place on [Street] came off the market after 94 days, and it looks like the price got adjusted twice during that run. You’re sitting on $250k of equity in that house, so there’s real room to work with on strategy. I’m curious what you think actually happened during those 94 days that kept it from moving.”
The second version isn’t a better-written script. It’s the same basic structure, said by someone who clearly did the work before the phone rang. That’s the whole difference, and it’s the part that data enrichment plays.
Turning the Call Into an AI-Assisted Follow-Up Sequence
A cold call that doesn’t convert on the first attempt is the norm, not the exception: most expired and FSBO appointments come from the third, fourth, or fifth touch, not the first. The same enriched data that improved the opening line should feed the follow-up: a text referencing the specific equity number, an email that mentions the actual days-on-market instead of a generic “just checking in,” a voicemail drop timed to when the listing would have hit its next price-reduction anniversary. Building that sequence out in your CRM is worth its own walkthrough. The point here is that enrichment isn’t a one-time input for the call; it’s the data layer the whole follow-up cadence should run on.
For the data side of AI in your business, see AI for Realtors: How to Use It for Market Analysis
FAQ
Does AI cold calling mean using an AI voice to make the calls?
Not usually, for residential agents. Most “AI cold calling” for realtors refers to AI-assisted script writing, list enrichment, and follow-up automation around a human-dialed call, not a fully automated AI voice caller, which carries its own TCPA consent requirements.
Is data enrichment legal to use on expired and FSBO leads?
Yes. Enrichment tools pull from public records, MLS data, and licensed data providers. The legal question isn’t the enrichment itself; it’s whether the resulting number is checked against the DNC registry before you call.
Do I need both a real estate-specific tool and a general enrichment platform like Clay?
Not to start. Begin with one real estate-specific source (REDX, PropStream, or BatchLeads) to build the list and pull the core data points. Layer in a general enrichment platform once that workflow is running and there’s a specific gap, like owner contact data on absentee or LLC-owned properties, that it isn’t filling.
Where to Start This Week
Pull this week’s expired list, run it through whichever real estate-specific tool is already in the stack, and add one enrichment field that isn’t already being pulled: equity position or days-on-market history are the highest-leverage starting points. Rewrite the first line of the call around that one data point before touching the rest of the script. That single change will do more for the call than any amount of rewriting the ask at the end.
