AI's Uncomfortable Truth: Are We Busy or Just Performing Busyness?
I'm often struck by how terrified everyone is of AI stealing their job, yet I rarely hear chatter about what scares me more: AI exposing the charade of busyness we've built our professional identities on.
Think about it. Your proptech platform can score risk factors across thousands of properties in seconds, but then what happens? That brilliance gets funneled into a Tuesday standing meeting where eight people spend 30 minutes debating whether to green-light something the algorithm already determined is a safe bet.
I worked with a commercial real estate firm last year that implemented predictive analytics for their portfolio management. The technology was flawless. Their processes? Pure organizational theater. The AI revealed that roughly 60% of their "decision points" were predetermined by existing policy, yet they maintained the same approval chains and meeting schedules they'd had since 2008.
We've gotten remarkably comfortable with artificial inefficiency. It feels safer somehow – as though the more people and meetings involved, the less blame any individual shoulders.
But inefficiency is expensive in ways beyond just time. While your competitors are streamlining decision flows, you're still "circling back" and "putting pins in things" that could have been resolved in minutes.
Maybe the real resistance to AI isn't fear of replacement but fear of exposure. What happens when the machine makes you confront how much of your workday is spent in performative productivity?
Totally agree that behavioral scoring’s getting sharper—especially with the explosion of AI/ML models crunching tenant data to rank risk like it’s Tinder for leases. But here's the thing: scoring isn’t the bottleneck. It's what happens after you've made the scoring decision that’s still stuck in the fax-and-clipboard era.
You can predict tenant churn with 92% accuracy, great. But if it still takes three departments, a spreadsheet, and a dozen emails to send a retention offer? You've just paved a Formula 1 track and asked a horse-drawn cart to race on it.
Let’s take maintenance workflows. Say an AI flags that Unit 306’s HVAC is about to fail based on usage patterns and weather data. Predictive models nailed it. But then the ticket pings an admin, who waits for a manual approval, who emails a vendor, who marks it “urgent,” and three days later, someone shows up. During that delay, the tenant’s already frozen, pissed, and sharing their love for your property on Google Reviews.
Amazon doesn’t just have great recommendation algorithms—they’ve got robots in warehouses, automated routing, and same-day delivery. They retooled the pipe, not just the brain.
In proptech, we’ve bolted smart risk engines onto architecture that still thinks it’s the '90s. So yes, you’ve got insight—but no way to act nimbly on it. That’s not innovation, that’s dressing up a dinosaur in a lab coat.
If proptech wants ROI, it has to stop treating process as an afterthought. Scoring is sexy, sure. But until the downstream operations are just as digitized—yes, end-to-end—you’re just model-rich and margin-poor.
You know what terrifies corporate America more than AI taking their jobs? AI revealing how much time they waste.
I was at a proptech conference last month where a CTO demonstrated their new mortgage processing AI. Everyone applauded the 80% time reduction, but nobody addressed the elephant in the room: if this system eliminates 30 hours of work per application, what were all those humans actually doing before?
The uncomfortable truth is that our organizations are designed around inefficiency. The weekly status meeting exists because our systems are too fragmented to provide real-time updates. The six approval signatures aren't about risk management—they're about distributing blame.
This is why digital transformation keeps failing. We digitize broken processes instead of reimagining them. It's like putting a Ferrari engine in a shopping cart and wondering why you can't take corners.
When we talk about "change management challenges," what we're really saying is: "Our organization has built identity and power structures around inefficiency, and we're terrified of what happens when that's exposed."
The proptech companies winning right now aren't just building better algorithms—they're designing entirely new operational models that make most of our current work rituals look like corporate theater.
What's your experience with this? Have you seen organizations confront this reality, or are they still digitizing their dysfunction?
Totally agree that behavioural scoring has come a long way — knowing when someone’s going cold on a lease renewal or which tenant will likely pay late is practically table stakes now. But here's the kicker: the ROI doesn’t just vanish because of "outdated operations." It dies because the infrastructure of most proptech isn't built to take action in real time.
We’ve become really good at insight and really bad at execution.
Think about it: You’ve got a machine learning model telling you a tenant shows high churn risk. Cool. But your leasing team still works off spreadsheets, the CRM requires manual updates, and legal needs three days to redline a renewal offer. By the time you respond, the tenant’s already ghosted and signed with the competition.
This disconnect isn’t just annoying — it’s profit leaking out.
Look at what Opendoor tried to do early on. Full-stack control. Buy, list, sell — all through their system. Say what you want about their unit economics, but operationally? They could act fast because they didn’t bolt new intelligence onto old plumbing. They rebuilt it. Most proptech startups today slap AI insights onto workflows that were designed for paper clipboards and fax machines. It’s like strapping a Tesla motor onto a horse carriage.
So yeah, behavioural models are smart, but until proptech stops treating operations as an afterthought and starts rethinking the full execution stack — contracts, approvals, compliance — margins will stay stuck in the 2000s.
The meeting problem isn't just about wasted time—it's about organizational theater that's suddenly transparent under AI's spotlight.
I was consulting with a proptech firm last month where they'd implemented AI for document analysis and contract review. Suddenly, their weekly 90-minute "status update" meetings felt bizarre. What used to take 15 people discussing ambiguities was now handled by an algorithm in minutes.
But here's the kicker: nobody wanted to cancel the meetings. They'd become social rituals, validation mechanisms, and—most importantly—proof that everyone was "working." The AI didn't threaten jobs directly; it threatened the elaborate dance of looking busy.
This gets to something deeper about work culture. We've built organizations where presence equals productivity, where meetings are currency in the attention economy. When AI strips away the need for collective head-scratching, what's left is often an uncomfortable void.
Think about how many proptech implementations fail not because the tech is bad, but because they expose organizational inefficiencies that nobody wants to acknowledge. The executive who championed six redundant approval steps doesn't want an algorithm pointing that out.
Maybe the real fear isn't replacement but revelation—that AI might show how much of our workday is spent on things that don't actually matter.
Exactly — we’ve built a telescope with behavioural scoring, but we’re still steering the ship with a compass from 1985.
The tech layer is not the bottleneck anymore. Look at underwriting. The models can now predict default risk with eerie precision. They can cut through noise like job title inflation or inflated self-reported income. But then what happens? That insight gets thrown into a clunky waterfall of forms, approvals, and legacy integrations that bottleneck the value. It’s the equivalent of calculating real-time traffic reroutes, then mailing someone the directions.
Take tenant onboarding. You’ve got AI models that can detect fraud better than a human underwriter ever could — parsing pay stub inconsistencies, reading metadata off PDFs, cross-checking names and leases. Fantastic. But then you still need someone in Ops to chase a “utility bill dated within the last 30 days” because of a checkbox in a process doc no one’s questioned in a decade. That’s not compliance — that’s inertia dressed up in process clothing.
The risk models are proving their worth, but the ROI falls off a cliff when processes weren't designed to match the slender signal-to-noise ratio of modern tools. We’ve gone from crystal balls to headlights, but we’re still driving in reverse because “that’s how the process works.”
Unless the back-end systems — both technical and organizational — get rearchitected around dynamic rather than static workflows, proptech ROI is going to keep looking like a gym membership: paid every month, but no one’s seeing real gains.
It’s not the risk-taking that’s dumb. It’s pretending the old processes were ever designed for the clarity we now have.
I think we're dangerously close to exposing corporate America's dirty little secret: most of our time isn't spent being productive, but performing productivity.
The AI revolution isn't threatening jobs so much as it's threatening our collective fiction about how we spend our working hours. When an algorithm can review 500 mortgage applications in the time it takes you to schedule your next alignment meeting, suddenly those two hours of "touching base" on Thursdays look suspiciously unnecessary.
I worked with a proptech firm last year that implemented a decent ML system for tenant screening. Suddenly their approval process went from 3 days to 20 minutes. The unexpected result? Complete chaos. Nobody knew what to do with the time saved. The weekly "application backlog" meeting still happened—it was just five people staring at each other with nothing to discuss.
This is the real transformation we're avoiding talking about. Not that AI will replace us, but that it will force us to confront how much of our work lives are structured around inefficiency rather than value creation.
The smart companies are already asking: if our AI can do in seconds what took human days, what could those humans be doing instead that actually matters?
Exactly — the intelligence is showing up in the scoring models, but it’s still hitting brick walls on the handoff.
Let’s take tenant underwriting as an example. You've got this slick ML model that can predict payment reliability with eerie accuracy, way beyond any human underwriter. But as soon as the model flags a strong applicant, what happens? A week of email chains, manual document verification, some poor analyst copy-pasting pay stubs into a 10-year-old CRM. That model might be smart, but it’s just feeding a clunky Rube Goldberg machine.
And it’s worse in commercial real estate. You’ve got platforms that can analyze a lease, assess tenant risk, even project ROI on unit splits. Yet the leasing process still runs on PDFs, legal bottlenecks, and a “Let me circle back with John from facilities”… who’s out until Tuesday. No one’s automating the fat parts of the workflow where the actual time and money disappear.
So here’s the paradox: Proptech’s cognitive layer is advanced, but its procedural layer is prehistoric. We’ve installed a Tesla brain into a horse and buggy.
The tragedy? The ROI doesn’t die because the tech is bad — it dies because no one’s reimagined the rails it's running on. There’s no reward for smarter risk if the approvals, deal structuring, and pipeline management are still glued together with Excel and prayers.
Until that changes, all we're doing is scoring faster in the first third of the funnel… and watching the rest of the deal bleed out in operations.
You know what's fascinating? When companies invest millions in AI to optimize processes, only to waste those efficiency gains in pointless status meetings where everyone takes turns saying "no updates from my end."
I watched this happen at a proptech firm last month. They'd built this incredible risk assessment engine that could do in seconds what took analysts days. Impressive tech. But then I sat in on their weekly "alignment" meeting—ten highly paid professionals spending 90 minutes sharing updates that could have been an email, while barely touching the critical decisions their new AI system had flagged.
It's like buying a Ferrari and then only driving it in school zones. The painful truth is that our meeting cultures evolved in an era of information scarcity, where gathering people was the only way to share knowledge. Now we're drowning in real-time data but still clinging to calendar rituals that make less sense by the day.
The truly disruptive companies aren't just implementing AI—they're reimagining their entire operational rhythm around it. Meetings become targeted decision forums rather than information exchanges. Updates happen asynchronously. And suddenly those efficiency gains translate to actual margin improvements rather than just filling newly available time with more meetings about meetings.
What's your experience with this? Have you seen organizations successfully shift their operational culture to match their technology investments?
Totally agree that behavioural scoring's come a long way—tenant screening today looks more like credit card fraud detection circa 2015 than the guesswork it used to be. But here’s the catch: scoring is the easy part.
The hard part—and where the drag really lives—is in everything that comes after. Take lease approvals. We’ve got models telling us with high accuracy that this tenant will pay on time and treat the property like a third child. Great. But then that decision gets funneled into a process that still relies on PDFs, manual approvals, and back-and-forth emails like it's 2004.
It’s like putting a Tesla engine in a horse-drawn carriage. Sure, the front pulls harder, but the whole system still turns like molasses.
And it’s not just about speed or elegance. These outdated operations create real financial drag. A property manager at scale might spend 40-50% of their time on repetitive admin work—chasing documents, logging into five different systems, matching payment receipts to units. That’s time not spent on optimizing portfolios or improving tenant experience, which is where the real ROI lives.
Worse, these clunky processes introduce friction that erodes the very signals behavioural scoring gives you. A high-scoring applicant waits five days to hear back, gets ghosted by an overwhelmed leasing agent, and takes a place down the block. Congrats, you just lost your best tenant because your backend runs like a DMV.
So yes, the models are getting smarter. But ROI doesn’t come from being clever—it comes from being end-to-end relentless about speed, simplicity, and scale. Until proptech starts treating process automation with the same excitement it gives to machine learning, we’re just scoring risk inside a Rube Goldberg machine.
You know what's fascinating? We've spent years optimizing algorithms to predict human behavior, but we're still sitting through two-hour meetings where nothing gets decided.
I was at a proptech conference last month where a CEO proudly showcased their new AI risk modeling system. Brilliant tech. Could predict default probability with remarkable accuracy. The room was impressed.
Then I asked how long their approval process takes once the AI makes its recommendation.
"About two weeks," he said.
Two weeks! Their algorithm works in milliseconds, but the human decision chain takes 336 hours. And that's not even counting the weekly "sync" meetings where everyone recites what they could have just written in an email.
This is the dirty secret of digital transformation. We've revolutionized data processing but left our organizational habits untouched. The AI exposes the meetings-for-meetings'-sake culture that's been hiding in plain sight.
Think about it - what happens when your AI platform instantly generates the insights that used to take your team days to compile? Suddenly those three update meetings you scheduled aren't justifiable anymore.
This isn't just about efficiency - it's about a reckoning with how we've structured work itself. The parts of your job that feel most productive (being busy, attending meetings) might actually be the least valuable.
Exactly. We’ve got algorithms that can tell you more about a tenant’s risk profile than their mother—but they’re still being processed by humans with PDFs and legacy CRMs that look like they were built for Y2K.
The irony? The smarter your risk models get, the more glaringly dumb your processes become. Because better insights are only valuable if you can act on them fast. And most property companies simply can’t.
Take onboarding. You’ve got models predicting payment delinquency with 80% accuracy at the application stage, but what happens next? Manual ID checks. Paper leases. Compliance bottlenecks. No auto-triggered workflows, just a team of people forwarding emails. It’s like dropping a rocket engine into a horse-drawn cart.
And let’s not forget maintenance ops. Predictive analytics can flag asset deterioration before it becomes a crisis—but if the work order still lives on a spreadsheet buried in Cheryl’s desktop folder, you’re back to reactive mode.
Proptech loves to talk about AI in the sexiest terms—computer vision for property inspections, GPT-powered lease abstractions—but too often these tools sit disconnected from the actual operational stack. That’s the ROI killer. Not the tech itself, but the orphaned workflows dragging everything down.
It’s not even a question of innovation at this point. It’s integration. Until the plumbing’s modernized, the intelligence is wasted.
You know what scares executives more than AI replacing jobs? AI exposing how much time we waste pretending to work.
I watched this play out at a proptech firm last month. They implemented an AI system to streamline underwriting—cut the process from 6 days to 4 hours. Amazing, right? Except then came the awkward realization that their twice-weekly, hour-long "workflow alignment" meetings suddenly made no sense. The entire pipeline those meetings were designed to manage had basically disappeared.
This isn't just about meetings, though. It's about the elaborate ecosystems of busywork we've built around inefficient processes. When AI strips those away, we're left staring at organizational structures that look increasingly absurd.
That mortgage company I mentioned? They didn't celebrate cutting underwriting time by 97%. Instead, middle managers scrambled to justify why they still needed their teams. The most telling moment was when someone suggested slowing down the AI system to "maintain operational stability."
Translation: "This efficiency is threatening our sense of importance."
The irony is painful. We've spent a decade optimizing for busyness—tracking activity metrics, holding status meetings, creating dashboards to monitor workflows that shouldn't exist in the first place. Now AI is calling our bluff.
What's your experience with this? Have you seen organizations confront this uncomfortable truth, or are they still in denial?
Totally agree that we’ve made leaps in behavioral scoring—tenant churn prediction, smart lease-to-income models, even using Amazon order patterns to detect co-living setups before they go rogue. That stuff’s sharp. But here’s the hangover: all that data wizardry is still getting fed into business processes that feel like they were built during the fax machine era.
Take leasing workflows. Some platforms can tell you a tenant has a 92% probability of default based on social signals, late-night logins, and five bounced subscriptions—but then what? That risk score hits a leasing rep’s inbox, who forwards it to finance, who copies a region manager, who wants to “get eyes on it.” By the time someone makes a decision, the tenant has signed elsewhere or ghosted entirely. The lag kills the value.
And don’t get me started on maintenance ops. We've got predictive models telling you an HVAC unit is likely to fail in 3 weeks. But if your system still requires a property manager to download a PDF, email a vendor, and wait for three quote responses, congrats—you've just created a beautifully modeled logjam.
This isn’t just a process issue—it’s a decisioning bottleneck. Proptech has obsessed over making information smarter, but not making actions faster. And ROI doesn’t come from smarter data alone; it comes from closing the loop between insight and outcome. That’s where I think the next real leverage lives.
This debate inspired the following article:
Smart risk, dumb processes: the real drag on proptech ROI. We’ve nailed behavioural scoring, but outdated operations are still killing margins.