Homework Before. Call After.
Expired listings have been the best agents' quietest goldmine for thirty years — the seller already proved they want to move. What changed is the size of the pile and the speed of the race. Article 16 says you cannot touch the seller before the agreement lapses, so you do the work before instead: AI predicts which listings will expire and builds the entire diagnosis in advance. Hour one, you arrive with a plan while forty other agents are still reading a script.
The full August 12 live session — the Death Watch, the Pre-Built Autopsy, and the First-Hour Play, start to finish
Four numbers that explain why this hour pays for itself
Expired listings are not a new idea. They have been the quiet goldmine of the best agents for thirty years, for one boring reason: the seller already proved they want to move. What changed in 2026 is the size of the pile and the speed of the race — half the market is sitting stale, and forty other agents get the same expiration notice you do at 8:00 AM.
So this session is not about finding expireds. Everyone can find them. It is about arriving with something nobody else has: a finished diagnosis of why that specific home did not sell, built by AI before the listing ever expired.
Read this before you build anything. It decides what the systems are allowed to do.
You may not solicit a seller whose listing is currently under an exclusive agreement with another broker. NAR's Code of Ethics Article 16 names this directly, and it specifically calls out telephone and personal solicitation of owners identified through MLS compilations as having current listings. An "about to expire" call list is exactly the thing the Article was written to prohibit.
There are only two narrow doors, and neither is a strategy you can run at volume:
So here is the line this whole session respects: before expiration you may research, analyse, and prepare as much as you like — because analysis is not solicitation. You simply do not contact the seller. The instant the listing lapses, the restriction lifts and you make your move. Homework before. Call after.
Your MLS may add its own rules about using MLS data for solicitation, and state DNC/DNE rules still apply to the call itself. Check both before you run this at scale.
The 8:00 AM race everyone else is running — and why it is the wrong race
Here is what happens to a listing the morning it expires. The notice hits every prospecting platform at once. Between forty and sixty agents pull the same record. The seller's phone starts ringing at 8:04 AM and does not stop. Every single caller opens with a version of the same sentence: "I noticed your home didn't sell, and I'd love the chance to…"
The seller has heard it eleven times before you dial. Speed is table stakes — it gets you into a queue, it does not get you the listing. And the sellers know the difference between a fast agent and a prepared one within about nine seconds.
What actually wins is documented: the agents who convert expireds research before reaching out, lead with empathy rather than a pitch, and bring a more specific plan than the seller had the first time. All three are things you can only do if the work was finished before the phone rang. That is what AI changes — not the speed of the call, but how much is already done when you make it.
AI ranks the active listings most likely to expire, and tells you when
This is an intelligence system, not an outreach system. Nobody gets contacted. You are building a weekly board that answers one question: which listings in my farm are going to die, and roughly when?
Seven signals do most of the work. Give AI the MLS export and the scoring rubric, and it produces the ranked board:
The output is a board sorted by predicted expiration date with a one-line cause of death for each. Run it every Monday. It compounds: by week four you are watching a pipeline, not a list.
Paste these into ChatGPT, Claude, or Gemini with your MLS export attached
You are a listing analyst for a real estate team in [MARKET]. I am attaching an export of ACTIVE listings in my farm area with these fields: address, list price, original list price, days on market, price change history, photo count, and public remarks. For each listing, score expiration risk 0–100 using: days on market vs. the market median of [X] days (weight 30), current price vs. my comp estimate (weight 25), number and timing of price cuts — a first cut later than day 45 is a red flag (weight 20), photo count and quality signals (weight 15), and description strength (weight 10). Return a table sorted by risk descending: address, risk score, days on market, percent over comps, cut history, and a single-sentence cause of death. Do not contact anyone; this is analysis only.
For each listing in the ranked table, estimate the likely listing-agreement expiration date. Use the original list date and the typical agreement length in [MARKET] (commonly 6 months; use 6 unless the data suggests otherwise). Flag anything expiring in the next 30 days as PRIORITY, 31–60 days as WATCH, and 61–90 days as HORIZON. Return address, original list date, projected expiration date, and tier.
Turn the PRIORITY and WATCH tiers into a one-page Monday briefing I can read in three minutes. For each property give me: the address, the projected expiration date, the single biggest reason it has not sold, and the one thing I would do differently — phrased as a plan, not a criticism. Keep each property to four lines. Rank by how winnable I am likely to find it.
Fill this in during the session with your own market numbers
The full diagnosis, assembled while the listing is still active — and nobody is contacted
This is the system that separates you from the other forty callers, and it is entirely legal to build because analysing a public listing is not soliciting a seller. You are writing a document, not making a call.
The autopsy has three parts, and each uses a genuinely different AI capability:
Those three become one client-ready page — "Why Your Home Didn't Sell" — plus a 90-day relaunch plan. You write it in advance, you sit on it, and you send it the hour the listing lapses. The seller has spent six months being told the market is the problem. You are the first person to hand them a specific, evidenced answer.
Photo, price, and copy — then the relaunch plan they can hold
You are a listing-marketing critic. I am attaching the photos from an active listing at [ADDRESS], listed at [PRICE] in [MARKET]. Analyse them as a buyer scrolling on a phone would. Tell me: 1) which photo is currently first and whether it is the strongest asset, 2) any photo that is dark, cluttered, badly angled, or actively hurting the listing, 3) rooms or features that appear to be missing entirely, 4) the ideal reordering of the first five images, and 5) the three specific reshoots that would change buyer response most. Be concrete and reference photos by number. This is internal analysis of a public listing — do not draft any outreach.
Here is the price history for [ADDRESS]: original list price [X] on [DATE], reductions on [DATES/AMOUNTS], current price [Y]. Here are six comparable sales I pulled: [PASTE COMPS]. Tell me: 1) what percent over the comp set it launched, 2) whether the first price cut came early enough to matter and what the research says about cut timing, 3) what the correct day-one price would have been, 4) what it is worth today given [X] days of market exposure, and 5) the one sentence that explains this to a seller without insulting the agent who took it.
Here is the public remarks section for [ADDRESS]: [PASTE]. And here are the property's three genuinely best features: [LIST]. Tell me: what the description led with versus what it should have led with, which of the three best features it buried or omitted, any language that is generic filler, and then rewrite it in 180 words leading with the strongest asset. Give me both versions side by side so I can show the contrast.
Using the photo, price, and copy analysis above, write a one-page document titled "Why Your Home Didn't Sell — [ADDRESS]". Structure: a two-sentence empathetic opening that does not blame the seller or attack the previous agent; three findings, each with the evidence and the fix; then a 90-day relaunch plan with weeks 1–2 (pricing and prep), weeks 3–4 (new photography and copy, relaunch), and weeks 5–12 (marketing cadence and checkpoints). Write it in my voice: [PASTE 200 WORDS OF YOUR OWN WRITING]. Professional, plain, confident, no hype.
Pick one real listing from your Death Watch board and build it live
What happens the hour it lapses — and the quieter play at 30, 60 and 90 days
The restriction lifts the moment the agreement expires. Now speed matters again — but you have something no other caller has, so your opening is completely different from theirs.
The 72-hour sequence:
And the play almost nobody runs: the 30/60/90-day expired. The stampede is over within a week. Sellers who did not relist immediately are still sellers — 90% try again within six months — and by day 45 you may be the only agent still talking to them. Same autopsy, zero competition. Put these on a recurring calendar rather than treating them as dead.
The delivery note, the call frame, and the follow-through
Write the short note that accompanies the "Why Your Home Didn't Sell" one-pager for [ADDRESS], sent the day the listing expired. Constraints: under 120 words, no pitch, no "I'd love the opportunity", no criticism of the previous agent. The note's only job is to make them open the attachment. Acknowledge that they have probably heard from a lot of agents today and that this is different because it is about their house, not about me. Sign-off should invite nothing more than reading it. My voice: [PASTE VOICE SAMPLE].
Based on the autopsy findings for [ADDRESS], give me a call frame for the follow-up conversation — not a word-for-word script. Include: the opening line that references the document rather than my services, the two questions that will get this specific seller talking about what frustrated them, the empathetic reframe when they blame the market, my one-sentence answer to "what would you do differently", and the specific close for a twenty-minute planning conversation. Anticipate the three objections most likely from a seller whose home just failed to sell after [X] days, and give me a response to each.
Write a three-touch sequence for a seller whose listing expired [30/60/90] days ago and who has not relisted. They are past the stampede and probably discouraged. Touch 1: a market update specific to their street with one line about what has changed since they delisted. Touch 2: a short case study of a relaunch that worked. Touch 3: a direct but low-pressure invitation. Each under 130 words, in my voice: [PASTE VOICE SAMPLE]. No urgency tactics and no fake scarcity.
Expirations are an echo of listings taken six months ago. That makes them forecastable.
Most agents treat expireds as random weather. They are not. A listing that expires in October was signed in April. Since listings peak in spring, expirations peak in the fall — the exact months most agents ease off prospecting.
Two consequences worth building around:
Run the projection each quarter. Knowing that thirty agreements in your farm lapse in the first two weeks of October is worth more than any lead list you can buy.
What to build first, and the four ways this quietly falls apart
Build in this order. Week 1, the Death Watch only — get the board running and trust the scores before you build anything on top. Week 2, one full autopsy on one property, start to finish. Week 3, the delivery note and call frame. Week 4, the 30/60/90 sequence and the quarterly calendar.
The four killers:
Four weeks from zero to a running pipeline
What one recovered listing is worth against the time this takes
Work your own numbers. The maths on expireds is unusually forgiving, because the seller has already proven intent — you are not creating demand, you are recovering it.
Sign it. Vague intentions do not survive a busy week.
"Nobody wakes up wanting to be an expired listing. Behind every one of them is a person who made plans, told their family, and watched six months go by while the phone didn't ring — and was told the whole time that the market was the problem. They don't need another agent promising to try harder. They need one person to hand them a specific, evidenced answer to the question nobody would answer. Do the homework before you're allowed to call. Then call, and bring the answer."
— Edmund Bogen