What to actually expect from an AI keynote speaker

An AI keynote speaker delivers a session on how artificial intelligence applies to a specific business context, usually 30 to 60 minutes, for a conference audience that is trying to figure out what to do next. The best ones are operators who currently run AI in their own work and can describe, in specific terms, what failed and what didn't. The weakest ones are trend-readers with polished decks and no working stack.

The problem with most AI keynotes right now

The supply of AI keynote speakers has grown faster than the supply of people who actually run AI in a real business.

That gap shows up in a predictable way. You get a 45-minute talk about the history of language models, a few brand-name case studies from McKinsey or Goldman Sachs, a lot of animated slide transitions, and a closing slide about "embracing the future." The audience leaves informed-ish and unable to do anything different on Monday morning.

For a general consumer audience, that's tolerable. For a room of real estate brokers, marketing directors, or brokerage operators, it lands flat. Those audiences are practical. They want to know: which tools, which workflows, what breaks, and how much does it cost to run. They know what a CRM is. They know what a lead magnet is. They do not need the definition of a large language model.

The bar for an AI keynote that actually lands with an operator audience is higher than it was two years ago. Most conference programmers are still calibrated to the 2023 standard.

What separates a practitioner keynote from a pundit keynote

The fastest diagnostic: ask the speaker what AI workflow they changed or killed in the last 90 days, and why.

A practitioner answers that question immediately. They name the tool, explain what it promised, explain where it broke down, and explain what they replaced it with. That's a speaker running AI in their actual work.

A pundit deflects. They'll tell you about a client who went through something similar. Or they'll name a trend. Or they'll answer the adjacent question about what AI is capable of rather than what they personally ran and abandoned.

This matters because the audience will ask versions of that same question from the floor. An operator audience wants to talk to someone who has been in the same problem space. They're looking for pattern recognition, not news.

I run AI across my consulting practice, my coaching programs, and Monstera Real Estate. That means I've had workflows that worked well, workflows that sounded promising and weren't, and workflows that needed three rebuilds before they produced something I'd actually use in a client engagement. When that comes up in a keynote, there's something concrete to discuss. The audience recognizes it because they've been in similar spots.

What a real estate audience specifically needs from an AI keynote

A realtor or brokerage operator audience has specific questions that a general AI keynote doesn't address:

• Which tools integrate with GoHighLevel, FollowUp Boss, or Kvcore without breaking the existing workflow?
• What does AI-assisted lead generation actually look like, operationally, day to day?
• Where does AI still need a human check before anything goes to a client?
• What's the minimum viable AI stack for a solo agent versus a team of ten?
• How do you train an AI tool on your market and your voice without spending weeks on it?

These questions don't show up in enterprise case studies about manufacturing or financial services. They need a speaker who has thought through the real estate context specifically, not adapted a generic AI talk by swapping in a few real estate examples.

The conference programmer's error I see most often is booking an excellent AI speaker from a tech or marketing background and assuming the real estate application is obvious. It's not always obvious to the audience. The translation work matters.

What formats actually work

The format question is underrated. For an AI keynote at a large real estate conference, a 45-to-60-minute talk with a tight structure works well: what the landscape actually looks like now, where operators are making decisions, what to try first, and what to avoid. Questions at the end.

For a smaller brokerage event, a half-day workshop tends to produce more lasting change. Agents can build a workflow during the session, not just watch one get described. They leave with something in their hands.

A 20-minute keynote at a dinner gala is not the place for AI depth. That format is for awareness and curiosity, not implementation. If a programmer books an AI speaker for that slot and expects agents to change their business practices from it, that's a format mismatch.

I'm direct about this in my own booking conversations. If the goal is genuine practice change across a brokerage team, the format needs to support that. A keynote that inspires but doesn't equip tends to produce a lot of nodding and not much follow-through.

What to check before you book

A few practical things to evaluate before booking any AI keynote speaker:

Ask for a recording. Not a reel. A full recording of a recent session. Reels are edited to look good. A full recording shows pacing, depth, how they handle live questions, and whether they go specific or stay abstract when pushed.

Check their public-facing work. A practitioner who runs AI in their own business has artifacts: a public blog with actual workflows described, social content that names tools, evidence that they're doing the thing rather than advising on the thing. If the only public presence is keynote stage photos, that's worth noting.

Ask what tools they currently run. Not "what tools do you recommend." What tools do you use, right now, in your own work. The answer should be specific: Claude for this, ChatGPT for that, GoHighLevel for this workflow. A speaker who hasn't thought through that answer probably hasn't built it.

Check whether their talk is built for your audience type. A keynote that worked for a marketing SaaS conference is not automatically the right talk for a real estate franchise convention. Ask how they adapt.

What I'd actually book, if I were the programmer

If I were programming an AI session for a real estate conference in 2026, I'd be looking for someone who does three things:

First, they show real operator-level work. Not case studies from clients they can't name. Their own work, in their own practice, with the specifics that come from having lived it.

Second, they understand the realtor's actual stack. CRMs, lead gen, follow-up sequences, listing content, client communication. The constraints a realtor operates under are specific. The speaker should know them.

Third, they give the audience something to do tomorrow. Not "explore AI" or "embrace the shift." A specific thing: the first workflow to build, the first tool to try, the first question to ask about their existing setup.

That's the keynote that generates the post-event email saying the session was the most useful one of the conference. It's not common. But it's the right bar.

FAQ

What should I look for in an AI keynote speaker? Look for a speaker who is actively using AI in a working business, not just talking about it. Ask what tools they run, what broke, and what they changed as a result. A practitioner can name the failure. A pundit pivots to a different anecdote.

What's the difference between an AI keynote and an AI keynote for realtors specifically? A general AI keynote covers broad trends and enterprise applications. A keynote built for real estate operators goes into the tools, workflows, and decisions that are specific to brokerage operations, including lead generation, CRM automation, listing content, and team management. The framing and the audience's questions are different.

How long is a typical AI keynote? Most conference keynotes run 30 to 60 minutes. A half-day workshop or breakout session can run 90 minutes to 3 hours with interactive components. The format should match your audience's goal because inspiration and skill-building are different experiences.

What topics should a good AI keynote for real estate cover? Practical tool selection, workflow design, lead gen automation, content production, CRM integration, and where AI still fails agents. Trend-only keynotes without operational depth tend to leave real estate operators with nothing to implement the next day.

How do I evaluate an AI keynote speaker before booking? Ask for a recording or transcript of a recent talk. Check whether they name specific tools, specific failures, and specific workflows. Review their public-facing work. A speaker running AI in their own practice will have artifacts to point to. Vague keynote reels are a signal.

Is an AI keynote speaker right for a smaller brokerage event? Often yes, if the session is structured as a workshop rather than a pure keynote. A 90-minute working session for a team of 20 agents tends to produce more usable takeaways than a 45-minute talk for 500 people. Match format to goal.

---

Emma Pace — strategic marketing consultant, AI coach for realtors, keynote speaker. Realtor at Monstera Real Estate. Builds AI-operated marketing systems at emmapace.ca.