The real estate community did not choose bad software. Bad software became the default.
That distinction matters more than it sounds, because the two explanations lead to completely different conclusions about what happens next. If agents chose bad software, the problem is the agents, and the fix is to convince them to choose better. If bad software became the default, the problem is the inertia of an industry that standardized on tools built for a way of working that no longer matches how agents actually live. The fix for that is not persuasion. The fix is building something that matches reality and letting the default shift.
I am in the second camp. I think the default shifted out from under the agents while they were busy doing their jobs, and almost nobody stopped to notice that the tools never shifted with them.
Here is what I mean.
Are real estate agents mobile workers?
Yes. Real estate agents are among the most mobile professionals in any industry. They run showings, take calls between appointments, answer emails from parking lots, return client texts at nine at night, and handle transaction paperwork from wherever they happen to be standing. The car is not where they commute to work. The car is where the work happens, and the largest line item in the typical agent's business budget is the cost of operating that vehicle.
Everyone who has spent time around this industry knows that real estate agents are mobile. They are not sitting behind a desk. They live in their cars. They run showings across town, take calls between appointments, answer emails from parking lots, return client texts at nine at night, and handle the steady stream of paperwork that every transaction generates from wherever they happen to be standing. A large share of an agent's working life happens in motion. Many of them eat most of their meals in the car between appointments. The car is not where they commute to work. The car is where the work happens.
The data backs this up in a way that is almost poetic. According to the National Association of REALTORS 2025 Member Profile, released in August 2025 and based on survey responses from 4,947 members collected in March 2025, the single largest business expense category for agents continues to be the cost of operating a vehicle. Not office rent. Not marketing. Not technology. The car. The largest line item in the typical agent's business budget is the thing that makes them mobile, because mobility is the job. When the biggest expense a profession carries is the vehicle it works out of, that is not a footnote about commuting costs. That is a description of where the work actually happens. The office, for a large share of agents, is wherever the car is parked at the moment a client needs something.
So you have an entire profession whose defining operational characteristic is that they work from their cars, on their phones, in motion, all day. And then you look at the software that profession was handed to do its transaction work, and the software assumes the exact opposite. The software assumes the agent is sitting at a desk, behind a keyboard, with two free hands, a large screen, and uninterrupted attention. The software assumes a working posture that describes almost none of the actual moments in which an agent needs to use it.
That mismatch is the whole story.
Why is most real estate software built for a desktop?
Most real estate transaction software was designed during the desktop era around the assumption that filling out forms is a seated, two-handed, full-attention activity at a desk. The industry standardized on those tools, the standard became the default, and the default persisted through inertia long after the working reality of agents had moved to mobile, phone-first, in-motion work. No one chose the desktop assumption deliberately. It was inherited and reproduced year after year because it was already there.
The transaction software most agents use was designed for a desktop era. It was built around the assumption that filling out forms is a seated, two-handed, full-attention activity that happens in an office. It was built for a version of the job that, for a lot of agents, barely exists anymore. The interface design, the input model, the entire interaction pattern assumes a stationary user. And the agents using it are anything but stationary.
I want to be precise about how this happened, because the word I keep coming back to is default, not choice.
Why has bad real estate software persisted for so long?
Bad software persists not because of bad decisions but because of inertia. Tools built during the desktop era became the industry standard, the standard became the default, and the default reproduced itself as every new agent and brokerage inherited it. The original design was reasonable for its time. The problem is the widening gap between the design's assumptions and the current mobile reality of the profession, a gap that grew every year while the core tools went unchanged.
No agent sat down and decided that the best possible tool for a mobile, phone-first, always-in-motion profession was a desktop forms application built around a keyboard. That decision was never made by anyone. What happened instead is that the tools got built during the desktop era, the industry standardized around them, the standard became the default, and the default persisted through sheer inertia long after the working reality of the profession had moved on. Every new agent who entered the field inherited the default. Every brokerage that onboarded them trained them on the default. The default reproduced itself, year after year, not because it was good but because it was there.
This is how almost every case of bad industry software happens. It is rarely a story of bad decisions. It is usually a story of decisions that were reasonable when they were made, frozen into a standard, and then carried forward unchanged while the world around them moved. The software was not built badly. It was built for a context that stopped being the context. The badness is not in the original design. The badness is in the gap between the design's assumptions and the current reality, a gap that widened every year while nobody updated the tools.
What does the NAR 2025 Technology Survey say about AI adoption among agents?
According to the 2025 NAR Technology Survey, released in September 2025, two out of three agents agree or strongly agree that their brokerage provides all the technology tools they need (38 percent agree, 29 percent strongly agree). However, only 17 percent of agents said AI has had a significantly positive impact on their business, while 46 percent reported no noticeable impact at all. The most widely used tools in the profession are eSignature and social media, which sit on top of the workflow rather than rebuilding the core transaction process.
There is a second layer to this that the NAR data makes visible. According to the NAR 2025 Technology Survey, released in September 2025, two out of three agents either agree or strongly agree that their brokerage provides all the technology tools they need. On the surface that sounds like contentment. But read it alongside the same survey's finding that only 17 percent of agents said AI has had a significant positive impact on their business, while 46 percent reported no noticeable impact at all. The tools are present. The tools are adopted. The tools are not producing the transformation the technology was supposed to deliver, because the tools that have been adopted are mostly the surface-level ones. ESignature. Social media. The survey found those are the most widely used tools in the profession. Useful tools. Tools that sit on top of the workflow rather than rebuilding it.
What has not changed is the core transaction workflow itself. The actual filling out of the actual forms. That is the part that was built in the desktop era and that is the part that has not been rebuilt for how agents actually work. The peripheral tools modernized. The core stayed frozen.
How could AI improve real estate transaction software?
AI makes it technically feasible for the first time to build transaction tools that match the actual working posture of a mobile agent. Instead of requiring an agent to sit at a keyboard, AI-powered tools can let agents complete transaction work in motion, by voice, between appointments. The core transaction workflow, the actual filling out of forms, is the part that was built for the desktop era and has not yet been rebuilt for how agents actually work.
And here is the part that makes this the right moment to talk about it. We have AI now. The thing that was impossible to build in the desktop era is possible now. For the first time, it is technically feasible to build transaction tools that match the actual working posture of the agent. Tools that work the way the agent works, in motion, by voice, between appointments, without requiring the agent to sit down at a keyboard and pretend they are back in 2007.
I am not going to pitch a specific product here. I am going to make a more basic point. The technology that would let us rebuild the core transaction workflow for a mobile profession now exists, and the only reason most agents are still using desktop-era tools is that the default has not shifted yet. The capability is here. The inertia is what is holding the line.
That is what needs to change. Not the agents. The default.
When a default has persisted for that long, shifting it takes more than a better product. It takes a product so obviously matched to how people actually work that the old way starts to look absurd by comparison. It takes the moment where an agent uses something built for their real working life and cannot understand why they spent years doing it the other way. Defaults do not shift through argument. They shift through a better option becoming undeniable.
The agents were never the problem. They are some of the hardest-working, most mobile, most adaptable professionals in any industry. They adapted to the desktop tools because the desktop tools were what existed. They will adapt to better tools faster than almost anyone expects, because the better tools will match what they were already doing anyway. The gap between how agents work and how their software assumes they work has been an open wound in this profession for close to two decades. The agents have been compensating for it the whole time, with laptops balanced on car seats and forms filled out in parking lots and workarounds layered on workarounds.
They deserve tools built for the job they actually have, not the job someone imagined they had during the desktop era.
We have AI. We have the capability to build for the mobile, in-motion, phone-first reality of the profession. The only question left is how long the old default holds before the better option makes it indefensible. My bet is that it holds for less time than most people in this industry expect, because the gap between the tools and the reality has gotten too wide to ignore, and because the agents themselves have been waiting for someone to close it.
The real estate community did not choose bad software. They inherited a default that stopped matching their reality. The fix is not to ask them to try harder with the old tools. The fix is to build them something that matches how they actually live and work, and to let the better option do what better options eventually do.
That is what every conversation about real estate technology should actually be about.
Judd Hoffman is CEO and Co-Founder of Ethica AI, building AI-powered tools for real estate transaction workflows.