AI Showing Follow-Up Messages for Real Estate Agents

After a showing, buyers often decide in hours whether they want a second look, more comps, or to move on. AI can turn your notes into a clear text or short email faster—but it can also invent feedback, over-promise next steps, or slip into fair-housing-risky language.

This brokerage-first workflow for US residential buyer’s agents covers fact sheets, SMS vs email, prompts for ChatGPT or Claude, an edit checklist, and high-level TCPA/privacy habits. It is not legal advice or a substitute for brokerage, MLS, or state association rules.

Why post-showing follow-up is different

A post-showing message is not a weekly nurture drip. It should reference this property and appointment, mirror feedback buyers actually gave, offer one clear next step (second showing, comps, lender check-in, or keep looking), and stay mobile-short. A text often wins same-day; email works when you need bullets or links.

Generic “Hope you loved the home!” drafts get ignored. Locked facts get replies.

Step 1: Build a showing fact sheet before you open AI

Treat AI as a rewriter, not a memory of the appointment. Capture:

Address or MLS # and showing date
Attendees by role only (e.g., “buyers”)—not family biography
True likes/concerns stated (layout, condition, price feel, commute)—no invented praise
Open questions (HOA, disclosures, seller credits)
One authorized next step and preferred channel if known

Never paste into consumer AI: IDs, SSNs, full loan files, or long CRM dumps. Summarize what the message needs. See AI privacy and compliance basics for realtors.

Step 2: Choose SMS vs short email

Channel	Best when	Length target	Watch-outs
SMS / CRM text	Same-day check-in; one clear CTA	Prefer under ~320 characters when possible	Consent and brokerage texting rules; avoid sensitive docs in SMS
Short email	Need bullets, listing links, or a mild paper trail	80–180 words + clear subject	Subject must match the body; no fake urgency

If unsure of channel, use the one they used to confirm the appointment—or ask once. For email subjects, see AI email subject lines for real estate.

Step 3: Prompt ChatGPT or Claude with hard constraints

Lock the facts and ban risky categories in the same prompt.

Prompt example (same-day SMS):

You are helping a US residential buyer’s agent draft a post-showing text.
Audience: the buyer clients who just toured the home.
Tone: professional, warm, concise—no hype, no pressure.

Facts (do not invent or embellish):
[paste address/MLS #, showing day, 2–4 true likes/concerns, open questions,
authorized next step, agent first name / brokerage placeholder]

Write ONE SMS draft:
- Under 320 characters if possible; max 500 characters
- Acknowledge the showing briefly
- Mirror their real feedback in plain language
- One clear next step + easy reply option (e.g., “Want me to pull comps?”)
- Sign with agent first name only

Hard constraints:
- No demographic, familial, religious, or “perfect for families” language
- No school/church/crime neighborhood stereotypes
- No claims that the home will appreciate or “won’t last”
- Do not add feedback, prices, or next steps not in my fact sheet
- Do not invent that they “loved” the home unless that word is in the facts
Return the text only.

Prompt example (short follow-up email):

Using ONLY these showing facts: [paste sheet].
Write a short email for buyer clients after today’s showing.
Include:
1) Subject line (under 60 characters, specific to the address—no clickbait)
2) Body: 100–160 words
Structure: thanks for touring → summary of what they liked/flagged (facts only)
→ one recommended next step → offer alternatives if they want to pass
→ professional close with agent name / phone / email placeholders.

Hard constraints:
- Fair-housing-safe: no family, school-as-persuasion, religion, or demographic hooks
- No fabricated urgency or investment promises
- No attachment of sensitive docs in the draft
- Flag anything that needs brokerage review
Return subject + body only.

Reject any draft that invents “you loved the kitchen,” adds a second showing you did not propose, or steers with school or family language you did not authorize.

Fair-housing-safe language (examples)

Safer (property / process)	Riskier (avoid)
“You liked the natural light and wanted clarity on the roof age.”	“Perfect for a young family like yours near the good schools.”
“Happy to request the seller’s disclosure packet next.”	“You’ll love the church community nearby.”
“Want me to line up a second look Saturday?”	“Safe neighborhood for kids.”
“If this one is a pass, I can send three other listings that match your price and commute notes.”	“Homes here don’t last—especially for buyers like you.”

When in doubt, stick to property features, process next steps, and preferences the clients already stated. For broader tool choice, see best AI tools for real estate agents (2026).

Brokerage / TCPA-aware habits (high-level, not legal advice)

Texting and email rules are office- and jurisdiction-specific. At a high level: use brokerage-approved channels; do not treat a portal lead phone number as SMS marketing consent; keep post-showing messages tied to an existing client relationship when that is your office’s framework; honor stop requests immediately; and use one AI thread per client so details do not bleed across deals.

This article does not interpret the TCPA or your state’s rules—ask your broker or counsel. Pair with AI privacy and compliance basics for realtors.

Edit checklist before you hit send

Address, date, likes, and concerns match your notes—not the model’s imagination.
Delete invented enthusiasm (“loved,” “perfect,” “won’t last”) unless the client said those words.
Confirm the CTA is something you can deliver soon.
Cut fair-housing-sensitive or neighborhood-stereotype language.
Confirm channel, length, and consent path match office policy.
You read the final version. AI drafts; you own the relationship.

A simple same-day workflow

Stage	Action
After the showing	Fill the fact sheet while details are fresh
Channel	SMS for one CTA; email if you need bullets or links
Draft	ChatGPT or Claude with locked facts + hard constraints
Edit	Human pass against notes; strip hype and risky language
Send	Through brokerage-approved CRM / dialer / email
Log	Note the outcome in your CRM

If an offer is next, keep “love letter” habits separate—see AI buyer offer cover letters. For listing-side cadence, compare AI just-listed and just-sold emails.

Quick checklist

Fact sheet filled from real showing notes
Channel matches client preference and brokerage rules
Prompt bans invented feedback and fair-housing-risky hooks
One clear next step only
Final message re-read by you before send
CRM updated after they reply

Bottom line

AI can turn a clean showing fact sheet into a same-day text or short email in minutes. It cannot remember the appointment, invent lawful consent, or keep you fair-housing safe if you feed it stereotypes. Lead with real notes, choose SMS or email on purpose, lock constraints in the prompt, and edit before send.

Official tools: ChatGPT and Claude. Confirm texting/email policies before you scale any post-showing sequence.