Meta quietly released something at the end of September that I think a lot of real estate agents are going to want to play with.
It’s called Meta Muse for Small Business, and while there are plenty of AI tools that can help you brainstorm a Reel or write a caption, what is massive about Muse is that it can connect directly to your Instagram professional account analytics, along with your Facebook Pages, Meta ad accounts and other business tools.
I connected mine and started playing with it, but instead of asking it to make more content, which is how I use AI most of the time, I wanted to see what would happen if I used it to understand all the content I had already made.
Eventually, we analyzed 1,133 of my Instagram posts. It helped me understand why my rolling views had fallen from roughly 4 million to 2.46 million, showed me that at least 27 of my 98 newest followers were real estate professionals, and identified patterns in the topics and formats that have historically performed best for me.
That was when I realized I had probably been using AI backward. If you have been consistently creating content for a few years, you already have your own data set of what works and what doesn’t, so before asking AI for another 10 Reel ideas, there may be more value in asking it to help you understand what your audience has already told you.
Here are five prompts I’m using to do that.
5 Meta Muse for Small Business prompts
1. Diagnose the problem before trying to fix it
Prompt:
“Analyze my Instagram performance over the past 90 days. Don’t give me content recommendations yet. First tell me where the biggest weakness is between reach, engagement, profile visits, follower growth and conversion, and show me the data that led you to that conclusion.”
A few months ago, my Instagram analytics showed roughly 4 million views in a 30-day window, and more recently that number had fallen to about 2.46 million, so naturally my first reaction was that something wasn’t working.
The story was a little different once I dug into the data. Some of my biggest posts, including videos with more than 1 million views, had simply aged out of the rolling reporting period. I would love to tell you I publish a million-view video every week, and I’m working on it, but I don’t.
My views were down, but roughly 84 percent were still coming from non-followers, and my content had reached more than 1.4 million accounts. The point wasn’t to ignore the decline; it was to understand what was causing it before changing a strategy that might not actually be broken.
2. Find out who is actually following you
Prompt:
“Analyze my most recent followers, and group them into the most common audience types you can identify. Don’t assume who my target audience is. Tell me what the data shows.”
We pulled roughly 100 of my newest followers, and of the 98 we could analyze, at least 27 had Realtor, real estate, broker or realty directly in their name or username.
Initially, I wondered whether that was a problem because I sell houses and still want consumers seeing my content, but last year we closed roughly $12 million in sales volume from referrals, and relationships with other agents have also created speaking opportunities, partnerships and other revenue. Trying to get rid of an audience already producing real business made very little sense.
Instead, I’m leaning into it while becoming more intentional about building consumer-focused lead pillars through my local Dana Point Daddy page and YouTube.
I’m also still making consumer-focused and humorous content on my main account because my past clients, database, sphere, friends and family are seeing and engaging with it. We’re still marketing directly to our database, hosting client events and doing the things that generate real estate business outside social media.
Your unexpected audience might be completely different from mine, but you probably shouldn’t decide whether the audience you’re building is good or bad until you know who is in it and what that audience is worth to your business.
3. Find the DNA of your best content
Prompt:
“Analyze my highest-performing posts from the last year and group them by topic, format and hook. Tell me what patterns repeat among my winners and which patterns I seem to be underusing.”
I spend an embarrassing amount of time studying other creators, saving their posts and trying to understand why something went viral, while apparently not spending nearly enough time studying the years of content I already had telling me what my own audience responds to.
We analyzed 447 of my carousel and single graphic posts and isolated 103 that used headline-style hooks. The winners repeatedly included income statistics that create sticker shock, local rankings people want to agree or disagree with, surprising changes in the Southern California housing market and California housing policies translated into something useful for homeowners.
This October, Inman turns its focus to AI and its rapid rise in real estate. During Artificial Intelligence Month, we’re digging into the startups shaking things up and the established players folding AI into their offerings.
I don’t need to make variations of those posts forever, but I can look for new topics that share the DNA of things my own audience has already told me it wants instead of constantly trying to recreate somebody else’s viral post.
4. Compare what you post most with what actually works
Prompt:
“Compare how frequently I publish each content format with how those formats actually perform. Where is my time allocation inconsistent with my results?”
At the time of the analysis, I had published 66 times over 28 days, and Reels were the format I was publishing most frequently even though they weren’t my best-performing format. Carousels were.
I’m not going to stop making Reels because different content has different jobs, but if I’m deciding where to spend another hour this week, I should probably know whether the format consuming most of my time is producing my best results.
5. Ask AI for ideas based on your winners, not everybody else’s
Prompt:
“Based only on the patterns you identified in my historically best-performing content, give me 10 new content ideas. For each one, tell me which past pattern it is based on and why you think it fits my audience.”
This is where the first four prompts become useful instead of just interesting because I can ask any AI tool for 10 content ideas and it will happily give them to me, but without context it is essentially making an educated guess.
Once it understands what my audience has historically responded to, I can ask a much better question: What should I make next based on what has already worked for me?
I’m still going to experiment, tell personal stories and make something weird because I think it’s funny, but now my own history gets a vote.
Before you ask AI to create, ask it to analyze
Meta Muse for Small Business has only been out for about a month, so I am sure I’m barely scratching the surface of what I can do with it, but this experiment has already changed the way I’m using AI.
If you’ve been consistently posting for a few years, you may already have hundreds or thousands of experiments sitting inside your Instagram, so before asking AI for another hook, caption or content idea, ask it to help you understand those experiments first.
For me, the most interesting thing AI has done for my content lately isn’t helping me create more of it. It’s helping me make better decisions about what I create next.
Krys Benyamein is a Realtor with Allstars Realty in Southern California. Get connected on Instagram or LinkedIn.
