Long considered an industry where intuition reigned and reliable forecasting seemed out of reach, commercial real estate is undergoing a data-driven transformation, quietly but fundamentally reshaping how investment strategies take shape. This evolution is unfolding not as theoretical speculation but in the hands-on work of executives like Ryan Severino, chief economist and head of research at BGO, whose experience was recently unpacked in a revealing conversation on the Altus Group’s podcast.
Far from simply importing Silicon Valley fashion into real estate, Severino’s approach combines macroeconomic analysis with what he describes as “bottom-up” granularity, the kind of hands-on number crunching that challenges long-standing industry skepticism about forecasting. “There’s this understanding in the real estate industry that the data is not very good, and can’t really be used for forecasting,” Severino observed, but he and his team are proving that belief outdated. Their collaboration with data scientists and advocates of artificial intelligence has, he says, “consistently demonstrated that variables like rent are actually much more forecastable with pretty good accuracy than I think the average person in our industry perceives it to be.”
From that blend of experience, analytics, and open debate comes practical guidance for commercial real estate firms looking to harness data science for more resilient investments:
Invest in Both People and Technology
For firms aiming to advance their forecasting, Severino emphasizes the importance of building teams with complementary skill sets. This means bringing together trained data scientists who can extract actionable insights from vast property and market data sets, alongside economists who understand macro trends shaping demand and supply. At BGO, collaborative meetings between these groups are routine. Both perspectives are needed to reveal market turning points and subtle shifts that might otherwise go unseen, such as unexpected spikes or lulls in local leasing activity.
Challenge the “Bad Data” Myth
Severino’s experience challenges a frequent refrain within the industry that real estate data is simply too messy to use for forecasting. In reality, the effectiveness of these data sets depends on skilled analysis and proper technological investment. By rigorously testing historical data sets against new machine learning models, the BGO team discovered that rental trends and market cycles could be forecast with more reliability than industry skeptics allowed. Severino argues that dismissing data quality can become a self-fulfilling barrier, and that meaningful progress comes from trusting—yet verifying—the value of available data through continual review, cross-checking and creative use of supplemental sources.
Make Research Actionable
Insights are only valuable when they directly support investment and portfolio choices. BGO’s process involves incorporating real-time model outputs into ongoing strategy sessions, allowing portfolio managers and acquisition teams to access scenario-based forecasts when evaluating assets. “We do splice them together into a coherent, cohesive worldview,” Severino notes, meaning that every actionable recommendation in strategy—whether to exit, hold, or deepen investment in a property—must be defensible by both the data and the macro context. Research shouldn’t remain a back-office function; it belongs at the deal table.
Leverage Advanced Analytics—Without Losing Judgment
While tools like AI and machine learning help find correlations or trends traditional analysis may miss, Severino cautions that models can’t replace all forms of expertise. His team uses advanced analytics to narrow the field—identifying which submarkets are likely to outperform average projections, for example—but then applies judgment informed by previous market cycles, experience with local dynamics and consultations with operational staff. This balance, Severino notes, is what “keeps us grounded when the models get ahead of themselves.”
Foster a Culture of Experimentation
At BGO, success has often hinged on openness to unconventional approaches. This includes pilot projects that test new data sets—like foot traffic analytics or credit card revenue—as signals for property performance, or building scenario models using external data not traditionally associated with CRE. By encouraging team members to bring forward new ideas, the organization cultivates a dynamic, learning-focused environment in which adaptation is prioritized over sticking with the status quo.
Synthesize Insights, Don’t Isolate Them
The best strategic decisions come from connecting data points across sources, timeframes, and teams. Severino describes a workflow in which data science outputs and economic modeling are shared regularly with investment, asset management and operational teams to ensure all information is weighed collectively. This integration helps the firm avoid the trap of overreacting to short-term—but ultimately insignificant—fluctuations and instead seek out robust, multi-dimensional signals for action.
Use Data as Both Opportunity Finder and Risk Detector
Finally, Severino points to cases where robust analytical models helped uncover emerging markets before they appeared on competitors’ radar—yielding deals that would have seemed bold or counterintuitive if based solely on gut feeling. At the same time, stress-testing portfolios with scenario analysis has allowed his team to spot early warning signs of downturns or underperformance and adjust accordingly before problems cascade. The result is both a broader view of what’s possible and a more consistent ability to minimize losses.
© Arc, All Rights Reserved. Request academic re-use from www.copyright.com. All other uses, submit a request to TMSalesOperations@arc-network.com. For more information visit Asset & Logo Licensing.