Energy

How to Evaluate AI Claims in Commercial Real Estate: A Practical Checklist

There is no shortage of AI promises in commercial real estate right now. Every software platform, energy management vendor, and building technology company seems to have added "AI-powered" to its marketing. The immediate challenge for real estate teams thus isn't finding AI solutions – there are plenty! – but figuring out which ones actually do what they claim.

That distinction matters more than ever. A recent survey found that among Canadian businesses using AI, 93% reported adoption but only 2% had seen measurable ROI at the time of the survey. None of this means AI doesn't work, but it does mean you ought to ask probing questions before you buy. What follows is a practical checklist for cutting through the noise.

1. Ask: What specific problem does this AI solve?

The first filter is the most important one. "AI-powered" is a description of technology, not a description of value. To get to the heart of the value, push prospective vendors to articulate precisely what their AI does, and what problems it solves for your team.

Good answers are specific: "Our AI detects when your chiller is running inefficiently and tells your engineer exactly what to adjust and why." Vague answers — "Our platform leverages AI to unlock the value of your building data" — are a signal to dig deeper.

Ask for the use case in plain language. If the vendor struggles to explain it without jargon, that's worth noting.

2. Ask: What data does it actually need to work?

AI is only as useful as the data feeding it. A common failure mode in building technology is deploying an AI tool before the underlying data is clean, normalized, or even fully connected. Before AI can optimize your building, it needs to understand what it's looking at — and that means well-structured, continuous, high-quality data from BMS systems, meters, and sensors.

Ask vendors these questions directly:

What data sources does the system require at minimum?
What happens to outputs when data is missing or inconsistent?
How does the platform handle portfolios with multiple BMS vendors or inconsistent point naming?
Does the system use a standardized ontology (such as Project Haystack or Brick Schema)?

A vendor who can't answer these questions clearly likely hasn't thought through the data problem carefully enough.

3. Ask: Is this actually AI — or automation dressed up as AI?

Not everything called "AI" involves machine learning, predictive modeling, or any form of intelligence. Rule-based automation, scheduled reporting, and basic threshold alerts are useful, but they're not AI, and they won't behave like AI when conditions fall outside their predefined rules.

The distinction matters because AI-based systems (like fault detection and diagnostics, or demand forecasting models) should improve with more data and adapt to changing building conditions, while rule-based systems won't.

Ask the vendor: "Does your system learn and adapt over time, or does it apply fixed logic?"

4. Ask: Where are the numbers coming from?

Energy savings of "up to 30%." ROI within six months. Carbon reduction at scale. Such claims appear frequently in vendor marketing, and some of them are certainly legitimate. But the range of conditions hiding behind "up to" can be enormous.

When you see a statistic, ask:

From which buildings? A single high-performing pilot or customer example may not be an indicative result.
Over what time period? Short windows can flatter.
Compared to what baseline? Without a well-defined baseline and a methodology for measurement and verification (M&V), savings figures are estimates at best.

Credible vendors will point you to specific case studies with verified savings, defined baselines, and ideally a range of examples.

5. Ask: Has this been validated in the field?

Proof-of-concept deployments are valuable starting points, but they don't tell you whether a system holds up across a diverse portfolio of buildings with different vintages, equipment types, and occupancy patterns.

Ask for:

References from buildings similar to yours in size, type, and geography
Evidence of performance across a portfolio, not just a marquee site
Data on how long the system took to deliver results post-deployment
Any independent or third-party validation of outcomes (like Noda’s study with Lawrence Berkeley National Laboratory)

6. Ask: What does implementation actually look like?

AI tools that require a 12-month implementation, a dedicated analyst team, or full BMS replacement aren't well-suited to most real estate organizations. The operational reality matters.

Questions to ask:

How long until the system produces its first actionable insight?
What does the integration process look like with our existing BMS and utility data?
Who on our team needs to be involved, and what's the ongoing time commitment?
What happens when buildings are onboarded or equipment is upgraded?

The answer should reflect a realistic picture of deployment, not just the best-case scenario. Be cautious of vendors who can only answer in ideal conditions.

7. Ask: How are results measured and reported?

AI that optimizes building performance in ways your team can't verify creates a trust problem — and an accountability problem. Any platform claiming ongoing energy, cost, or carbon savings should have a transparent, auditable way to measure them.

Look for:

A stated and defensible M&V methodology
Attribution of savings to specific actions, not just ambient building conditions
Reporting that your sustainability team and finance team can both understand
Compatibility with your own reporting requirements

If a vendor can't clearly explain how they calculate savings, you'll have a hard time defending those numbers internally– let alone to investors and other external stakeholders.

8. Ask: What does support look like after go-live?

Technology platforms for building operations still largely hinge on human operators in buildings, which means enterprise portfolio deployments entail a considerable degree of organizational behavior change for the all-important value-creating, savings-generating work to get done. Thus, understanding the expertise and support that sits behind the software matters, especially during the first months of deployment.

Ask:

Is there a team of building systems experts and engineers who support implementation and review outputs?
What does onboarding and training look like?
What's the escalation path when the system flags an issue that requires judgment?
What's the support SLA?

The best AI platforms for buildings combine automated intelligence with expert human oversight and support. Automation handles the scale; experienced engineers handle the nuance and ensure your teams on the ground get the training and enablement they need to make full use of the technology.

A Quick Reference Checklist

Before committing to any AI platform for building operations or energy management, run through these questions:

What specific problem does this AI solve — in plain language?
What data does it need, and what happens when that data is missing or messy?
Is this adaptive AI or rules-based automation?
Where do the savings figures come from, and are they independently verified?
Has this been proven across a portfolio comparable to mine?
What does implementation actually require from my team?
How are ongoing results measured and attributed?
What is the platform support strategy after go-live?

The Bottom Line

AI in building operations has genuine, proven value — in fault detection, energy optimization, demand management, and carbon reduction. But the gap between what AI vendors promise and what they consistently deliver in the field remains wide enough to reward scrutiny.

About Noda

Noda is a data and analytics company on a mission to make every building smarter, more efficient, and more sustainable. Recently ranked in the top 10 tech companies leading the charge on climate action, its AI-powered suite of products surface unique insights that empower real estate teams to reduce costs, decrease time spent on routine work, and find and act on opportunities to save energy and carbon. Discover how Noda's solutions can unlock the potential of your assets and accelerate the transition to net zero. Visit us at noda.ai to learn more.