The short answer is yes. Operators are doing true data-driven maintenance (DDM) today. Productivity is up. The preventative and predictive maintenance that the industry has been promising, and charging for, for the better part of a decade is finally being delivered.
Here's the catch. So is the Paper-Mache version. The kind that looks like data-driven maintenance from a distance, photographs well in a board pack, and quietly collapses the moment you push on it.
The job for everyone running buildings right now is to tell the difference.
Why DDM has failed until now
Before we get to what real DDM looks like, it's worth being honest about why it has taken this long.
There are two reasons.
First, the technology let down the people who run buildings. Done properly, DDM creates more opportunities for savings and improvement. Many more. But for years, the technology didn't balance the equation by reducing time and making the work easier. The goal should never have been more work. The goal is more results, for the least time, with the most joy. The industry kept missing that.
Second, clients didn't make it clear that they expected this capability. Opportunities were identified but not addressed. Teams needed a small push from the client to drive improvement, and that push didn't come. Clients are respected. They get attention from the teams on the ground. A bit of leadership goes a long way, and for too long it was missing.
There's a third dynamic worth naming, even though it isn't really a third reason. The defences are coming down. In the past, collaborators would use whatever excuse they could to avoid change or accountability for performance. What we're seeing now is that the best of them don't fear either. They're great at their job already. They can see how this kind of data-driven intelligence, combined with their training and experience, is a much better place to be than where the industry used to sit.
That shift matters. It's what makes the next chapter possible.
The goal was never more work. The goal is more results, for the least time, with the most joy.
So what's the difference between true DDM and Paper-Mache DDM?
True DDM has to do four things. Each one has a Paper-Mache lookalike sitting right next to it. If you don't know the difference, you'll buy into the wrong one.
1. Collect and process data that becomes trusted intelligence
True DDM doesn't just hand you what the system already has. It creates new, trusted information and intelligence. Data an experienced engineer can rely on to make a decision.
🚨 Paper-Mache Alert: The "rubbish in, rubbish out" problem is amplified in AI environments where it's easy to be fooled into thinking more data is the answer. Just pour everything into a big data lake and the AI will sort it out. It won't. Even before agentic AI, this approach was proven not to work in countless failed Digital Twin rollouts. In an agentic world, guesses and hallucinations accumulate even faster. This isn't a technology problem. It's a trust and context problem.
You can disagree with the recommendation. You should never have to disagree about the data or the context that's relevant to you. A measure of performance in one geography or one type of space can be a great result in one place and a bad result in another. You don't just need data. You need context.
2. Drive maintenance automation
Data should drive automation. Just like bank reconciliation in accounting. Experienced professionals shouldn't be spending their time on things that can be automated, and the transaction cost of those activities should be falling from hundreds or thousands of dollars to cents.
🚨 Paper-Mache Alert: Automating standard checks like DA19 looks like DDM. It helps, but if that's all you do, it doesn't address the real problem. How do I collect the data I need to know what's actually going on?
Most DA19 automation projects are a compromise built on a false assumption. Checks are spaced out based on the hours available to the people doing them. The number of checks is limited by technology and by compliance. You check at the frequency the standard sets out, a frequency designed to enable the most important checks in a previously feasible timeframe. The result is essentially a spot check, often working off data that's months or years old, in the hope of catching something.
That's not what teams want. Teams want to know what the highest priority is now. The assumption that live information was impossible is completely redundant. We have buildings running billions of checks and balances a year, generating the information and intelligence to actually drive maintenance.
3. Drive maintenance augmentation
Some things can't be automated. That doesn't mean they can't be data-driven.
Real DDM does two things at once. It removes steps that free up expensive resources' time, and it puts the right information in front of those people so they can apply their expertise quickly. Like a doctor walking into a consultation already holding the lab results.
This is where real DDM earns its keep: fault diagnosis and root cause analysis, solution analysis, workflow routing to the best accountable person on the fastest path to fix, trending, benchmarking, capital health assessments, and reporting.
🚨 Paper-Mache Alert: "Actionable insights." Don't give busy people more problems without solutions or context. It's easy to surface an anomaly, a spike, or an energy trend that looks high. That isn't driving maintenance. The engineer still has to go on an investigation journey to determine the cause, and whether they're already aware of it and dealing with it. That's not augmentation but homework.
4. Don't isolate data inside a single building
Context inside a building matters. Context across buildings matters more.
A piece of equipment failing 5% of the time doesn't sound like much. But what if that's 40% above the benchmark? True DDM should drive both improved standards (getting better) and clear best practices, benchmarked against other buildings and equipment, not just inside your portfolio but across the industry. Done well, it can also make smart decisions about over-servicing, which is the conversation the industry almost never has but should.
🚨 Paper-Mache Alert: Using the term "best practice" without being able to show what it is in detail, how it's tracked, and how it's benchmarked. If the vendor can't show you the working, it isn't best practice. It's marketing.
How to tell which one you're buying
If you're evaluating a DDM solution right now, here are the questions to ask. None of them are unreasonable. All of them are uncomfortable for a Paper-Mache vendor.
- Are you giving me new trusted intelligence, or just relabelling data I already have?
- Can you show me where your automation reduces the transaction cost of maintenance to cents, not where it just spaces out checks I already do?
- When you surface an anomaly, do you also tell me the cause, the solution, who should fix it, and whether we're already on it?
- What is your benchmark for "best practice," how is it tracked, and how is my building performing against it?
- Where does your platform reduce my team's time and increase my results, simultaneously?
A real DDM platform answers all five quickly. A Paper-Mache one gets vague around question two and starts changing the subject by question four.
The industry has to do better, and it can
The data-driven maintenance the industry has been promising for years is finally real. That's the good news. The less comfortable news is that a lot of what's being sold as DDM still isn't.
Operators have lived through too many cycles of being told the technology will save them, only to inherit more work and less clarity. That's not acceptable anymore. The bar is higher. The capability exists. The best people in the industry are ready for it.
The job now is to make sure we hold ourselves, and every partner in the supply chain, to the real version. Not the one that looks the part, but the one that actually moves the needle.
That distinction is everything.
.avif)
