Broker owner Deb Siefkin writes that as AI makes CMA faster and cleaner, it is exposing gaps that most agents were not trained to address.
We have more data than ever before. CMA can be generated in minutes. AI can take comps, analyze trends, and explain pricing in confident, perfectly audible terms.
Participate in the INMAN Intel Index Survey
That should make pricing easier. But for most of us, it quietly makes pricing difficult.
It’s not just the tools that have changed. it is a role
For a long time, the pricing exercise looked like building a CMA. Collect comps. Build the range. Provide recommendations. It took time, and that time created the impression that the process itself was expertise.
The same process can now be completed in seconds. And when that effort disappears, so does the idea that creating a CMA is a job.
What remains is what most of us are not explicitly trained to do: make decisions under uncertainty.
I was reminded of this recently while working with the family of an elderly woman who was no longer able to live at home full-time. Coming in, I made assumptions that I think most agents would make. I thought I wanted to move out of my house as soon as possible and move on.
So when I spoke to them, I presented them with three pricing options. Solid market price. and a more advanced, more patient approach. I expected them to land in the middle.
They chose the highest one.
When I asked why, the answer was simple. They owned the house outright. they weren’t in a hurry. They didn’t need the money on the schedule. What they wanted was the full value of the assets without having to liquidate them.
The data didn’t tell us that. Comp didn’t tell me that. None of the tools I ran told me anything like that. It came from sitting at the table and asking the right questions.
CMA does not generate prices. it produces information
What happens next is up to interpretation. You need to decide which comps are most important, how the market is reacting to comps, what buyers are actually comparing the home to, and how seller priorities will shape the acceptable outcome.
Nothing is solved by better data. That’s the job.
This is why pricing can often seem difficult in data-rich environments. More information does not eliminate uncertainty. The number of possible interpretations increases. AI will accelerate it. This will give you a faster answer, but it doesn’t resolve what that answer means in your particular situation with a particular seller and a particular buyer on the other side.
Most pricing conversations are still based on observation. This is what the comp says: This is how the market moves. This is where I think we should go.
These are useful inputs. But they are not decisions.
That’s where the risk comes in. It’s not the data itself, but the confidence that comes with it. When the AI presents you with a set of ranked comps, a price range, and a clear narrative, it gives the impression that the conclusion is already there. However, the most important variable is still missing.
real job
Which will the buyer choose? What outcome is the seller trying to achieve? What trade-offs are acceptable?
Those questions don’t exist in the data. They live in conversation.
Agents who struggle with AI are not those who refuse to use it. They will be the ones who use it without realizing where its usefulness ends. AI is very good at organizing input. No meaning can be assigned to them in a particular human context. And that difference changes the work.
The job is no longer about generating analysis. The job is to frame decisions.
That means defining what’s important to you before you look at the data, identifying what buyers are comparing before choosing comps, and aligning your pricing strategy to the results sellers actually want, not just what the data suggests.
Once that structure is in place, AI becomes a powerful tool.
Otherwise, AI will be persuasive. And with unstructured persuasion, you’ll lose listings, misprice your home, and destroy the trust you’ve built.
The industry has spent years trying to get better tools. Faster tools. cleaner tool. Smarter tools. But tools don’t fix undefined thinking. They amplify it.
CMA is no longer a job. Here’s the decision.
That family taught me something I already knew but hadn’t named clearly enough.
Prices were not available in the data. It was in a conversation that almost never happened.
Deb Siefkin is a practicing broker and founder of RightSize Realty Associates. Connect with Deb on LinkedIn and Instagram.
