48%
of broker-site inquiries got no response at all when researchers posed as consumers.
Dated, and we say so. It's still the primary source behind every half-of-leads-go-unanswered claim.
WAV Group, January 2014 ↗Free class · Clients and follow-up
The client loop breaks in two places: the inquiry nobody answers and the past client nobody calls. This class shows you how to draft every client message from one context kit, so the reply is ready in seconds and year one after closing is already on your calendar.
Part 1
A funnel ends at closing. Your business doesn't. Past clients come back and send the next client, so the journey loops through eight stages:
Lay the published numbers over those stages and two of them stand out. Stage 2, response: the inquiry comes in and nobody answers. Stage 7, post-close: the deal closes and the agent disappears.
48%
of broker-site inquiries got no response at all when researchers posed as consumers.
Dated, and we say so. It's still the primary source behind every half-of-leads-go-unanswered claim.
WAV Group, January 2014 ↗88%
of buyers say they'd use their agent again or recommend them. Only 17% used an agent they'd worked with before.
Among sellers, 28% used an agent they'd worked with before.
NAR 2025 Generational Trends ↗39%
of buyers say they haven't recommended their agent to anyone since the purchase.
NAR 2025 Generational Trends ↗In NAR's 2025 Generational Trends report, 40% of buyers were referred to their agent by a friend, neighbor, or relative, and 17% used an agent they'd worked with before. 66% of sellers found their agent through a referral or prior use. Every cold channel consumers reported hiring from was in single digits.
The same report found 75% of buyers interviewed only one agent, and 81% of sellers contacted only one agent before choosing. There's rarely a bake-off. And sellers now stay in their homes a record 11 years, so the relationship has to survive a decade to earn the next listing.
The intent is there. The system isn't. That's a memory problem, not a charm problem, and memory is the part you can automate.
Part 2
You don't need eight new tools. You need one document of durable facts that you paste at the top of every chat. After that, every ask is three lines long.
Your name, brokerage, the market you really work, and what you actually sell. Two plain sentences.
Three messages you've actually sent clients, redacted. Unpolished is better. Voice comes from samples, not adjectives.
The local detail AI can't look up: the neighborhoods, the road projects, the quirks you explain every week.
How fast you reply, which day clients hear from you, what you promise. These become the promises your drafts keep.
The inspector, the roofer, the HVAC company you'd actually send a client to. This turns a reminder into a useful referral.
Wins you can back up. If this section is thin, leave it thin. An empty proof section is honest. An invented one is a problem.
No kit yet?
Paste three real sent messages and one line about who you are. That's enough voice and identity to draft with today. Fill in the rest as you go. When an ask needs a fact the kit doesn't have, add it in one line.
Redact before you paste
No client names, addresses, phone numbers, emails, or financials ever go into the chat. Replace them with placeholders like [BUYER] or [STREET]. You add the real details when you send.
Part 3
Draft your first touch for the channel you actually work, then have the model pre-build five instant replies. Read each one against your pasted messages. If it doesn't sound like you, name the sample it should match and run it again. Then install them where you reply: text replacement on your phone, email templates, or your CRM's quick replies. Answering in five minutes becomes filling in one bracket from a parking lot.
First touch and five instant replies
[Paste your context kit above this line] Draft my first-touch message for [CHANNEL: sphere note, open house follow-up, website form, or cold email]. Match my voice samples. Under 150 words. Lead with something useful to them, not an ask. List anything I must verify before sending. Then draft five instant-reply templates, one for each lead door: portal inquiry, sign call, referral intro, open house, website form. Same rules. Give each one a single [BRACKET] for me to personalize.
Three parts, every week, every active client. The third part matters most, because "what I need from you" turns an FYI into motion. Save the ask in your chat tool and put a recurring 15-minute block on your calendar, same day every week. The update still goes out when nothing moved: here's why, and here's the date it moves next.
Weekly client update
[Paste your context kit above this line] [Paste this week's deal facts, redacted: no names, addresses, numbers, or financials] Draft this week's client update. My voice. Under 150 words. Three parts: 1. What happened 2. What happens next 3. What I need from you If nothing moved, say so and give the date it moves next. Flag anything I must verify before sending.
Pick one real past client and draft all five messages in one sitting. Before you close the chat, create four dated calendar entries or CRM tasks, each naming the draft to refresh and send. The reminder fires, you refresh the draft, you send it. The system remembers for a decade. You stay on the send button. Repeat it for years two and beyond from the same kit.
Post-close year one
[Paste your context kit above this line] [Paste the past client's redacted close facts: close month, home type, anything they cared about] Draft my post-close year one: - a closing-anniversary note - two seasonal maintenance reminders that use my vendor bench - one market update for their neighborhood My voice. Each under 120 words. No ask in any of them. Then draft the month-12 referral ask: one honest, low-pressure message that makes referring me easy.
Score five out of five or redraft. Redrafting is one sentence: tell the model which question failed. If a draft says rates are moving or a home won't last and you can't back it with MLS data, a lender, or public record, it fails.
Under CAN-SPAM: an honest header and subject, identified as an ad, your postal address, and a working opt-out you honor within 10 business days.
Under the TCPA, you need prior written consent before the first automated text. Consent can be revoked at any time.
The FCC says AI-generated voice calls without consent are illegal. Never let an AI voice agent cold-call for you.
One-at-a-time calls and texts you type yourself sit under lighter rules, but Do Not Call screening and the Telemarketing Sales Rule still apply.
That's the legal reason email is the safest cold channel and automated texting is the riskiest. When someone opts out, they come off the list that day.
Free with a student account
Create a free student portal account to get the workbook, prompts, templates, and downloads for every class.
Create my free account →Already a student? The same link signs you in.
This method doesn't do that, on purpose. AI drafts from your context kit, you verify the flagged facts, and you send from your own account. Automated texts need prior written consent under the TCPA, and the FCC says AI-generated voice calls without consent are illegal. Automate the reminder. Keep the send human.
Put year one on the calendar before you forget: a closing-anniversary note, two seasonal maintenance reminders that name real vendors you trust, and one neighborhood market update, none of them asking for anything. Then send one honest, low-pressure referral ask around month twelve. Create the four calendar entries the same day you draft them.
As fast as you honestly can. In NAR's 2025 Generational Trends report, 75% of buyers interviewed only one agent and 81% of sellers contacted only one agent before choosing, so the first good answer often decides it. Pre-built instant replies in your own voice turn a fast answer into filling in one bracket.
Yes, with rails. CAN-SPAM is an opt-out law. A compliant commercial email needs an honest header and subject line, identification as an ad, your valid postal address, and a working opt-out you honor within 10 business days. Texts work the opposite way: automated texts need prior written consent before the first message.
Three real messages you've sent clients, one active deal's facts, and one past client's close month and home type, all redacted with no names, addresses, phone numbers, emails, or financials. Any foundation model works, such as ChatGPT, Claude, or Gemini. The free tier is enough.
Clients and follow-up
Turn a Google Sheet into a relationship memory, let AI organize and draft, and reach out only with a real, verified reason.
Clients and follow-up
Be your client's first AI voice, stress-test AI answers together, and gate every claim before it reaches a deal.
Clients and follow-up
Find the one useful next action for a contact, draft only when permission and facts support it, and record what really happened.
Library
The whole library of recorded classes, by topic.
Research sources
Where buyers and sellers found their agent (Exhibits 4-4 and 7-1), how many agents they interviewed or contacted (4-6 and 7-2), whether buyers would use their agent again (4-14), and how often they have recommended them (4-12).
The record 11-year typical tenure before selling.
Researchers posed as consumers and inquired through broker sites. Dated, and the primary source for the no-response figure.
The requirements for commercial email, including the postal address and opt-out rules.
Consent rules for automated calls and texts, including AI-generated voice calls.
The Do Not Call and telemarketing rules that still apply to manual outreach.
Product access, limits, local listing rules, platform policies, and laws can change. Re-check the linked authority before important or public work.
Keep building
AI Acceleration: Real Estate is a private, self-directed curriculum for practical AI work across content, communication, lead follow-up, CRM, client research, operations, and relationship management. Build it. Review it. Keep what works.
Explore the Skool community ↗Follow Ryan + AI Acceleration
No ranking, revenue, closing, or time-saving promise. The classes teach the work and the review systems around it.