In the first post of this series, we talked about why data, not ambition, is the real constraint on global energy efficiency, and why portfolio-wide visibility is the foundation for meaningful progress. In the second, we explored how owners can move beyond one-off retrofits toward a practical, scalable framework for deciding where to invest first across diverse portfolios.
Between those two steps sits an uncomfortable reality for many portfolios.
Even with good data and a clear prioritization framework, the same pattern repeats: the “best” buildings go first. The Class A assets. The flagships. The buildings with strong teams, mature systems, readily available data, and clean business cases.
That’s understandable. Those buildings are easier to reach, more visible, and often where leadership expects action. But they are rarely where the largest untapped opportunity sits.
The tier no one wants on their slide deck
Most portfolios have a long tail of buildings that quietly fall into what I think of as Tier 2—or sometimes Tier 3 (Class C) assets. These are smaller, older, or less digitally mature buildings. They may lack a connected BMS or submetering. Often, there isn’t a highly experienced facilities manager dedicated to the site.
From a portfolio perspective, these buildings are frustrating. The data is thinner. ROI calculations come with more uncertainty. Even when the theoretical savings look compelling, the risk feels higher.
This leads to a reasonable (but dangerous) question: If the savings are really there, why hasn’t anyone done this already?
Most of the time, the answer isn’t bad math. It’s that the operating reality makes the math hard to trust, and a deep energy audit isn’t a priority.
Why “it depends” still applies
In the previous blog, I explained why there is no universal retrofit playbook for existing buildings. That challenge becomes even more acute in this part of the portfolio.
A hospital in Boston behaves differently from one in Phoenix. Two office buildings of the same size and age can perform very differently based on controls, operations, and climate. Rules of thumb help, but they rarely survive contact with real operations.
That uncertainty is exactly why these buildings get deferred. Without reliable data or the staff capacity to constantly tune systems, even reasonable upgrades can feel like bets you don’t want to make.
Paradoxically, these are the buildings where technology can have the greatest impact.
Energy intelligence isn’t just for the best buildings
When people hear “energy intelligence” or “AI in buildings,” they often picture advanced use cases: predictive maintenance, digital twins, long-term forecasting. While those are potential applications, they’re also easiest to deploy in buildings that are already digitally mature.
That’s not where the real leverage is.
For smaller and mid-sized buildings, energy intelligence is about something much more practical: continuous oversight without requiring more human time. Pattern detection that surfaces issues no one can look for. Validated operational guidance that adapts to this building, not a generic model.
This isn’t flashy AI. It’s machine learning (ML) applied to improve how all types of buildings behave in the real world, especially when they weren’t originally designed for efficiency.
A problem no one would have found manually
Consider a large university campus in the northeastern United States. No comfort complaints. No year-over-year spikes in energy use. The building wasn’t targeted for a retrofit.
After deploying a lightweight sensor network and applying machine-learning-based monitoring, something stood out: there was almost no difference between weekday and weekend energy use.
The root cause was simple but deeply buried. The incoming air temperature setpoint was off by roughly 20 degrees, forcing the building to reheat air year-round, even in summer. The setting had likely been wrong for years. Nothing had changed, so nothing triggered an alarm.
An experienced facilities manager could have found it with time and focus, but neither was available. Fixing that single issue reduced energy use by more than 20%, without a major retrofit or disruption.
AI isn’t replacing people; it’s supporting them
In this part of the portfolio, AI isn’t about automation for its own sake. It’s about support.
It means facilities teams don’t have to check every building manually. It means anomalies are flagged even when nothing obviously “breaks.” It means lessons learned in one building don’t have to be relearned across dozens more.
That frees up human expertise for what actually matters: strategic decisions, planned investments, and real emergencies, instead of chasing inefficiencies.
Why this matters
From a climate perspective, emissions don’t care whether a building is Class A or Class C. And from a portfolio perspective, it’s often the sheer number of these less glamorous buildings that makes them decisive.
If we want emissions reductions at the speed and scale required, we need solutions that work where time, staffing, and certainty are limited, and not just where conditions are ideal.
Energy intelligence puts better tools into harder places. And very often, that’s where the biggest gains have been waiting all along.
For portfolios where data gaps and uncertainty make retrofit planning challenging, the Impact Building Calculator can offer a practical next step. Use it to model retrofit scenarios using energy and carbon conservation measures (ECCMs), based on actual building performance when available, or simulation-based estimates when it’s not. Compare energy, carbon, and cost savings across investment levels to build confidence in early-stage planning, whether the building is Class A or Class C. To explore the tool with your next retrofit project, contact us.
Add a comment