Hey there, folks. Happy Tuesday. There's a question I keep coming back to lately: if brokerages are pouring millions of dollars into AI tools for their agents, why are so many agents still doing their best work with a free ChatGPT account? New survey data from Inman Intel just made that gap official — agents are seeing "significant" productivity gains from AI, but the wins are mostly coming from third-party tools, not the AI features their brokerages built and paid for. Meanwhile, broker-owner Deb Siefkin wrote a sharp piece on a platform called Ownli that's exposing an uncomfortable truth about what clients have been paying agents for all along. And in a more practical vein, Berkshire Hathaway HomeServices' Jimmy Burgess just dropped five specific AI prompts that every agent should be using right now. The theme this week is intentionality — the agents who win aren't the ones using more AI. They're the ones using it on the right things. Let's dig in.
1. Agents Are Getting "Significant" AI Gains — But Not From the Tools Their Brokerage Provides
Inman Intel dropped a fresh survey analysis yesterday that should make every brokerage executive uncomfortable. The headline finding from Daniel Houston's reporting: real estate agents who report "significant" productivity gains from AI overwhelmingly credit third-party tools — ChatGPT, Claude, Gemini, Copilot — not the AI features built into their brokerage's tech stack. The brokerage-funded tools that were supposed to be the differentiator? Largely sitting on the sidelines.
The pattern has been consistent for months. Earlier Inman Intel research found that agents who pay for "reasoning model" subscriptions (the higher-tier paid versions of these third-party tools) were three times as likely to report AI made them "significantly more productive" — 33% versus 10% among free-tier or other users. The new May survey reinforces that gap. The bright spot for brokerages: new internal AI tools rolling out in 2026 are starting to push beyond simple text generation into image and video creation, database management, and back-office document processing — the parts of the workflow where third-party chatbots can't easily compete. Houston's reporting suggests that's where the next round of brokerage AI will live or die.
Why It Matters: This is a wake-up call for every brokerage executive who thought they were going to win the AI war by building proprietary tools. The truth is that the consumer AI race has moved so fast — and reasoning models are improving so quickly — that purpose-built real estate AI is struggling to keep up with general-purpose tools agents can buy themselves for $20 a month. For agents, the practical takeaway is simple: if you're not paying for a reasoning model (ChatGPT Plus, Claude Pro, Gemini Advanced), you're leaving most of the productivity gains on the table. For brokerage leaders, the question is whether your AI investment is actually moving the needle for agents — or whether you're paying for tools your agents don't use because they prefer the chatbots they already trust. Survey your agents. Find out what they actually use. Then decide whether your platform is differentiating or duplicating.
2. Ownli Isn't Competing With Agents — It's Exposing What Clients Were Actually Paying For
Broker-owner Deb Siefkin wrote one of the sharpest Inman op-eds I've read this year, and it's worth reading twice. The piece is technically about Ownli — a consumer-facing platform that launched nationwide in March, offering buyers and sellers software to manage their own transactions and bypass traditional commission structures. But Siefkin's real argument isn't about Ownli. It's about what Ownli's existence forces every agent to confront.
Her thesis: most of what agents do — coordination, form routing, document chasing, status updates, MLS lookups, "just following up to see if you got my last email" — is now software. Consumers have been paying commission for processes that have been digitized in every other industry. They didn't realize it because the agent was the visible evidence of something getting done — but that "something" was never what they were actually buying. Ownli's launch matters not because it'll succeed at scale, but because it forces the conversation. The agents whose clients describe them as "responsive, organized, and easy to work with" aren't being complimented in 2026 — they're being commoditized. The agents who can articulate what their work actually produces — pricing strategy, negotiation outcomes, risk management, judgment under uncertainty — are not the agents Ownli is built to replace. The ones who can't are exactly the ones who should be paying attention now, while there's still time to change what they're selling.
Why It Matters: This pairs perfectly with the Intel survey above. AI is eating the routine tasks. Ownli (and platforms like it) is monetizing the fact that consumers have figured out coordination is a feature, not a value. The agents who survive the next five years are the ones who can answer one question cleanly: when a client describes what you did for them, what do they actually say? If the answer is "responsiveness, organization, communication" — those are commodities now. If the answer is "she saw something in the comps that nobody else did" or "he negotiated us $40K off after the inspection" or "she protected us from a deal that would have blown up" — that's value. Real, durable, expensive value. The good news is that bar isn't actually that high to clear. The bad news is most agents have never had to clear it explicitly. Time to start.
3. Five AI Prompts Every Real Estate Agent Should Be Using Right Now
Berkshire Hathaway HomeServices Chief Coaching Officer Jimmy Burgess published a practical Inman piece this weekend with five specific AI prompts every agent should have in their back pocket. The thesis is consistent with the rest of this issue: the agents getting the most out of AI aren't asking random questions. They're using specific, intentional prompts tied to specific business outcomes.
Here's the framework. Prompt 1 — AI Discoverability: Ask AI to analyze how it currently describes you and the agents recommended for your area, then suggest how to improve your bio and content so AI recommends you. This is GEO (Generative Engine Optimization) in practice. Prompt 2 — Listing Strategy: Have AI act as a marketing strategist to analyze your listing, audience, and recent comparable sales, then build a custom marketing plan including content, pricing strategy, and follow-up sequencing. Prompt 3 — Storytelling: Use AI to turn an MLS description into an emotional story that helps buyers envision the lifestyle of the home — formatted for Instagram carousel, a one-minute video script, and a LinkedIn blog post. Prompt 4 — Buyer Consultation Prep: Have AI build a personalized buyer consultation script based on the client's stage, budget, timeline, and concerns. Prompt 5 — Database Reactivation: Ask AI to draft personalized re-engagement messages for sleeping contacts based on their last interaction and the current market context. The common thread: every prompt is built around a specific business outcome, not vague exploration.
Why It Matters: This is the practical follow-up to everything above. If the data shows agents are getting big gains from third-party AI tools, and the market is exposing which parts of the job are commoditized, then the obvious question is: what should you actually do tomorrow morning? Burgess's answer is to stop "experimenting" with AI and start running specific plays. The AI Discoverability prompt is especially important — most agents have never actually checked how ChatGPT or Claude describes them when a consumer asks "who's a good agent in [my area]." Take 10 minutes today and run that prompt. The results will tell you whether you exist in the AI search ecosystem or you're flying completely blind. Start there, and work down the list. By the end of the month, you'll have five new AI workflows running and a much clearer sense of where AI moves your needle versus where it's just noise.
That's a wrap on this one, folks. The takeaway this week is consistent across all three stories: stop treating AI like a casual experiment. Agents are getting real productivity wins — but only the ones using the right tools on the right tasks. Brokerages are realizing they can't out-build ChatGPT on basic content tasks, so the next wave of internal AI is going to live or die on workflow integration. And platforms like Ownli are reminding everyone that the work clients have been paying for is changing. The agents who run intentional plays — like the five prompts above — and who can articulate their real value when asked are going to do just fine. The ones who treat AI like a magic wand are going to keep being disappointed. See y'all Friday.