This was day two of our month long live series on aged expired listings. On day one I showed you why I skip the expireds that just came off the market. This session was about the next step: taking the raw data from your MLS and cleaning it down to the short list of people actually worth your time.
The first half is the process: the MLS filters, the export, and the AI prompt that scrubs the list. The second half is me loading those leads into LeadDeck and calling them live. I've left out every homeowner name, address and number from the calls, but the steps are all the same ones you'll see on the screen.
Why aged expireds are worth the work
Aged expired listings are some of the best leads in real estate. These are people who already signed an agreement with an agent and put their home on the market. In my words on the stream, "They've already told the world that they want to sell." They just didn't have the right agent or the right plan to solve their problem.
The other big reason: when you pull this data from your own MLS, you're not buying the same list every other agent in your market can buy. You're building your own opportunities from data you already pay for.
Quality over quantity
A lot of expired listing services will tell you that you need hundreds of leads and hundreds of dials an hour. I don't buy that. "We want quality over quantity."
The goal is to spend about an hour working around 10 good cards, and go deep on each one, instead of burning through a giant list.
The MLS filters I use
I'm in Bright MLS, but these filters work in any MLS. You may just have to hunt for where each one lives.
- Pick your area first. On this stream I pulled a township near me.
- Remove new construction, and leave out bank owned and HUD properties.
- Pick your structure types. I chose detached, twin or semi detached, end of row townhouse and interior row townhouse.
- Set a minimum price. I used $100,000 and up.
- Choose expired and canceled, from 32 to 480 days back. That reaches back into some of the gold from 2025.
- Select all the data fields and export. In Bright MLS, the standard export has what you need.
Someone asked why I select everything in the export. Because the AI needs all of the data. If you give it part of the data and ask it to fill in the rest from the internet, sometimes it's good and sometimes it's not. "AI is really good when you give it all of the data."
Scrubbing the list with the MLS money prompt
Next I upload that export into an AI tool along with our aged expired MLS money prompt. It should work in ChatGPT, Gemini, Claude, or whatever AI you like best. It takes a few minutes to go through every row.
The prompt filters the list down to the clean candidates, the ones that are still real opportunities. The list I pulled the day before, from a rural county north of me, came back with 35 good candidates.
Then I asked the AI one more question to see what that list is really worth. Total up the original list price of every property, divide by 10, and multiply by 2.5% and 3%. That gives you a range of potential GCI if just one out of every 10 of those owners raises their hand. It helps you see the value of the hour you're about to spend on those cards. Version 2 of the prompt will do that math for you automatically.
The full prompt is free in the new member bonuses in our Skool community, along with the MLS filter checklist and a video walkthrough.
Start with what you have: pocket buyers
I also told a story about an agent in our Accelerator program who has a few pocket buyers. These are retail buyers looking for something specific in one or two zip codes, and she hasn't found the right home for them yet.
So we're going to use the same filters to find aged expireds that could fit those buyers. Best case, an owner says yes, we're still open to offers, and she has several interested buyers. Worst case, she can pivot to an institutional offer, or ask for the chance to get the home back on the open market.
If you're not sure where to start, just start with the type of properties you'd want to sell.
Loading the list into LeadDeck
To get contact info, I imported the scrubbed file into LeadDeck with skip tracing. In a minute or two, it matched 33 of the 35 and turned them into cards ready to call, text and email. I assigned them to the imported sellers sequence, which has the scripts.
On the next day's stream I promised to show how to get that contact data for free if you'd rather do it yourself. I use LeadDeck because I'd rather spend my time on conversations.
Live calls: going deeper on each card
My opener isn't about relisting the home. In our 1 Deal A Week blueprint we start with buyers. "Why do we start with buyers? Because without a buyer, we're not getting paid and neither is the seller." So I call to ask if they're still open to accepting offers.
Here's how I work each card:
- Call and leave a personal voicemail, then tie it to another channel. On a landline I follow up by email, since I can't text a landline.
- Hide bad numbers so I don't waste time on them next time.
- Respect the Do Not Call list. I only call a number on the DNC when I can see the owner is publicly marketing that number, for example on a Google Business Profile. LeadDeck won't let you tap to send a pre-populated text to a DNC number.
- Enhance the skip trace when the data is thin. That's common in rural areas.
- Send direct mail when there's no working phone or email. One tap adds them to a three piece mail campaign that says what I would have said on the phone.
As I put it, "I would much rather spend more time on one card and actually close a loop." Going deeper on each card means you need fewer cards.
When the answer is no, ask one more question
One owner told me the home had already sold off market. I didn't stop there. I asked if there were any other properties they'd consider an offer on, and what they planned to do with the proceeds. Those were all no's this time, and that's fine.
A no isn't always the end of the story. On day one, an owner who was no longer selling the home I called about told me about two others he does want to sell. Sometimes "no just opens the door to the better opportunity."
The hand raiser
One owner said yes, he'd still be open to an offer. The home has a tenant, so it could be a turnkey rental for the right investor. He also mentioned another property and a just renovated home he's deciding whether to flip or keep.
Here's what I did right away:
- Promised preliminary numbers that afternoon or the next morning, and confirmed his best email.
- Saved him as a contact in my phone with notes, so he's a pocket seller and a pocket buyer I can find fast.
- Pushed him to my CRM with one tap. I use BoldTrail, and LeadDeck works with any CRM that has a Zapier or API connection.
- Tapped to get preliminary offer options on his property, so I can come back to him with a real offer from a real buyer.
Key takeaways
- Aged expireds are great leads because they've already shown they want to sell.
- Pull your own data from your MLS, export every field, and let an AI prompt scrub it down.
- Work fewer cards and go deeper on each one.
- Lead with an offer from a buyer, not a pitch to relist.
- Tie every voicemail to an email, a text or a mailer, and respect the DNC.
- When someone says no, ask about other properties before you hang up.