AI is about to make every agent look more professional. That’s the problem

Andrew Fogliato uploaded a thumbnail and a transcript to ChatGPT’s new image generator during a live episode of their weekly sales and marketing show. On the first attempt, it produced a near-professional thumbnail that kept both hosts’ faces intact, used a real hook pulled from the transcript and included design elements that made sense.

No designer. No brief. Seconds.

“Do you know someone who’s never met you before, who, given access to the internet the way that this does and given only what you’ve told them, would have ever come this close to the mark on the first shot?” Taylor Hack asked. “Not a chance.”

When everyone looks professional, professional stops meaning anything

The same tool is available to every agent. If a non-designer can generate a near-professional thumbnail in seconds, the noise is about to get significantly louder. The agents who have been standing out because their marketing looked better than everyone else’s are going to find that advantage disappearing quickly.

Hack framed it directly. Everything AI touches gets replicated and multiplied fast. The casual users will use it for quick output. The serious operators will use it to go deeper than they could before. Those two groups are already diverging, and the gap will widen.

“The way that we are about to replicate the noise now visually – the noise is going to be so loud,” Hack said.

The differentiator that remains is the one AI cannot generate: a specific point of view, original insight and content that only you could have produced. Fogliato described this as the difference between commodity content and non-commodity content. Commodity content is anything anyone could write. Seven tips for buying your first home. Non-commodity content is specific, sourced and tied to a real experience, like: Three things I learned helping a client buy in Sherwood Park and how we saved $18,000 on the deal.

“Google is now working to prioritize that style of content,” Fogliato said. “The specific stuff over the generic stuff.”

How serious operators use AI differently

Most agents who use AI use it as a shortcut. They ask it to write a listing description, take the output and send it. The problem with that approach is that the output is exactly what every other agent using the same tool with the same level of effort is producing.

Fogliato described a better approach: instead of asking AI to write for you, ask it to interview you. What questions would you need to ask me to make this better? Then answer those questions and let the output reflect your actual knowledge and experience.

“What’s going to make it the best is original insight,” he said. “Nothing will outperform that.”

The same principle applies to ongoing campaigns. Most agents build a drip sequence once and leave it. The better move is to let it run, then feed the performance data back into the same conversation. Open rates, click-through rates, reply rates. Now the AI is not guessing at what works. It is improving based on what actually happened with a specific audience.

Fogliato gave the same advice on YouTube content. Export your analytics, feed them in and let the data shape which video ideas are worth pursuing. The brainstorm becomes faster and the output becomes more accurate to what that particular audience actually responds to.

The authenticity problem AI cannot solve

One of the clearest points in the episode was about the gap between what AI produces on technical metrics and what it produces on authenticity.

AI-generated emails tend to be grammatically correct, scannable and well-structured. On paper, they score well. In practice, clients recognize them immediately and the response they produce is closer to toilet paper than trust.

“An email written by AI: grammatically way better than yours. Authenticity so low people look at it like toilet paper,” Hack said. “The only thing that can stop that is to actually provide a disclaimer and say, here is an explanation I approve.”

The same problem is coming for listing photos and virtual staging. Hack noted that buyers are already suspicious. They can tell when a couch was dropped into a room digitally. They notice when the scale is off. The uncanny valley effect that makes AI video unsettling applies equally to AI-touched images once people know what to look for.

The agents who will benefit from the new image tools are the ones using them to iterate faster toward something authentic, not to replace the authentic thing entirely.

The live SEO experiment starting now

At the end of the episode, Fogliato and Hack committed to a public experiment: ranking Taylor’s website for a specific Edmonton real estate keyword using AI, step by step, on future episodes.

The structure is simple. Pick a keyword. Run the analysis in Claude. Build the page. Track the results. Do all of it in front of the audience so anyone watching can replicate the process.

The experiment grew out of a live demo Fogliato ran during the show, in which he asked Claude to analyze the first page of Google results for a competitive Edmonton real estate search term, then build a page designed to outrank them for Taylor’s ideal client profile. The output was specific, structured around Google’s current ranking priorities and substantially better than what most agents have on their sites.

“Even that version, without any additional prompting, would probably already be sniffing the first page within a couple of months,” Fogliato said.

The next step is picking the keyword publicly and beginning the process on air so the audience can follow along and apply the same approach to their own markets.

Join the conversation live

Each week, The Leads Are Sh*t brings working agents together to dissect what’s actually working in real estate right now.

Be part of the live recording, get your questions answered and join the after-show conversation.

👉 Register for the next live session.

Related Posts