Optimizing Place Performance with Simulation

Enterprises sometimes fear that AI will replace architects or service designers. The reality is the opposite.

Places - workplaces, hotels, hospitals, airport lounges, schools - are complex ecosystems to support peoples’ experiences, diverse roles and behaviors that aligns with organizational goals. Traditional design methods struggle to capture this complexity, while critical data streams live in silos, owned by different departments and not unified in a way that could inform planning and design effectively. Until recently, the tools to normalize and simulate this data at scale simply didn’t exist.

AI offers a new path. By linking enterprise data with simulation, organizations can test new services, forecast occupancy, test planning scenarios, and evaluate policies before they are put in place - giving planning, design, and operations teams stronger inputs while enabling leaders to plan with greater accuracy, adaptability, and efficiency.

Forecast Beyond Dashboards

Dashboards summarize the past. Simulation predicts the future. Simulation gives you the flexibility to adapt quickly when the data and assumptions change. By creating virtual “agents” that reflect employee or customer roles and schedules, enterprises can test hundreds of design options and forecast trade-offs before investing in construction.

This transforms portfolio planning from reactive adjustments into proactive optimization. In experiments, simulation helped participants uncover hidden design flaws—like circulation bottlenecks—early enough to iterate solutions. For corporate real estate, that reduces costly reconfigurations and accelerates time-to-value.

Learn What Teams Really Need

One of the most striking findings is that simulations help uncover incongruities between intent and actual performance, revealing the features and design strategies that are most likely to deliver the highest impact. Designers often expect certain layouts to spark collaboration, only to see different behaviors emerge.

Simulation exposes these mismatches, enabling organizations to learn how teams interact with space and uncover the hidden factors driving workplace performance. For executives, this translates to better alignment between place-based investment and business outcomes - reducing attrition and improving engagement by shaping environments around lived behavior, not assumptions.

Empower, Not Replace, Design Partners

Enterprises sometimes fear that AI will replace architects or service designers. The reality is the opposite. Simulation strengthens partnerships by giving designers richer, evidence-based parameters to work with.

Instead of handing architects vague requirements, organizations can provide data-driven insights into space ratios, encounter probabilities, and user satisfaction benchmarks. This elevates the dialogue from “what do we think might work?” to “how do we design around what we know actually happens?”—a strategic advantage in both cost and design quality. This approach powers critical shifts in design and planning strategy:

From target-based to range-based decisions – Beyond “we need 1,000 desks” to “we need so support x capacity for y on-site role mixes of between 850–1,050.”
From averages to distributions - Not “average attendance is 60%” but “on most days we x distribution of team-type utilization, a variation of 40-75% depending on team mix, with a X% chance of 90%+.”
From fixed benchmarks to dynamic models – Not “industry standard is 250 RSF per person” but “our own data shows a space and seat-type demand curve unique to our on-site teams.”
From one-way decisions to scenario planning – Beyond “three anchor days seems about right” to “what if anchor days increase or decrease, what are the right anchor day behaviors, and how does cross-team collaboration impact team attendance?”
From anecdotal feelings to evidence-based parameters - Not “executive X thinks we need more conference rooms” but “simulation shows room shortages at peak hours unless ratios or policies shift.”
From post-occupancy reaction to pre-investment forecasting - Not “we’ll survey after the move” but “we can stress-test layouts before we build.”
From anecdotal to quantified collaboration - Not “people bump into each other more here” but “encounter probability increases 22% with this layout.”
From cost vs. experience trade-offs to optimized balance curves - Not “either save money or support employees” but “optimize for both at the service level we choose.”

Empower, Not Replace, Design Partners

From target-based to range-based decisions – Beyond “we need 1,000 desks” to “we need so support x capacity for y on-site role mixes of between 850–1,050.”
From averages to distributions - Not “average attendance is 60%” but “on most days we x distribution of team-type utilization, a variation of 40-75% depending on team mix, with a X% chance of 90%+.”
From fixed benchmarks to dynamic models – Not “industry standard is 250 RSF per person” but “our own data shows a space and seat-type demand curve unique to our on-site teams.”
From one-way decisions to scenario planning – Beyond “three anchor days seems about right” to “what if anchor days increase or decrease, what are the right anchor day behaviors, and how does cross-team collaboration impact team attendance?”
From anecdotal feelings to evidence-based parameters - Not “executive X thinks we need more conference rooms” but “simulation shows room shortages at peak hours unless ratios or policies shift.”
From post-occupancy reaction to pre-investment forecasting - Not “we’ll survey after the move” but “we can stress-test layouts before we build.”
From anecdotal to quantified collaboration - Not “people bump into each other more here” but “encounter probability increases 22% with this layout.”
From cost vs. experience trade-offs to optimized balance curves - Not “either save money or support employees” but “optimize for both at the service level we choose.”

Author

Richard Tyson

The Future of the Intelligent Workplace ›