About GenslerOnWork

GenslerOnWork examines the modern workplace and how design can help us become more engaged and productive as we earn our livings.

Search GenslerOn
Work Topics
Connect with Us
« Stay Nimble and Expect Change: A Conversation About the Financial Services Industry | Main | The Changing Nature of Facilities Management »

Design Value Through the Lens of Big Data

Photo courtesy of Gensler

I have an unorthodox background for someone who works in the architecture and design industry. As a real estate analyst and statistician, one side pursuit over recent years has been using empirical analysis to obtain a better understanding of what has, up to this point, been widely understood but often hard to measure objectively: the business and economic value of great design. In this and subsequent blogs I will discuss the process through which I am exploring these relationships as well as initial findings and questions raised throughout my investigations.

Why am I doing this? The era of Big Data is upon us, and as designers we, along with other professionals, need quantifiable insights to understand how and where design creates value. Gensler’s history in this respect is strong – our workplace surveys and Workplace Performance Index (WPI) tool connect workplace design with employee satisfaction and business performance, building on ongoing research conducted by our research teams that began with our 2006 and 2008 workplace surveys.

Inspired by these efforts and the opportunity to use data analysis to inform design solutions, I embarked upon a parallel but distinct stream of research focused on commercial office buildings. Three years ago I began an investigation into a “premium” (above market rent) for office buildings designed by world-class architecture firms (as defined by Architect magazine’s top 50 lists). Through grants sponsored by Gensler’s internal research program and a partnership with my former professor of real estate finance, Chris Redfearn of USC’s Price School of Public Policy, I conducted research to identify a rent premium for multi-tenant office buildings based on national real estate data from the CoStar database.

This research is not without precedent. Strides have been made by those in the academic community to model rent premiums in commercial real estate, and I have been encouraged to give it a try. Previous efforts by professors Quigley, Kok & Eichholtz to identify a premium for LEED-certified buildings informed the goals and methodology of this project in particular, and our initial findings put that work in new light.

The research began with what is called a “null hypothesis” — a statement that is counter to the effect you believe is true, and that you will seek to disprove through analysis. For this endeavor, the hypothesis I sought to invalidate was that controlling for numerous factors — building submarket, class (A/B/C), age, size, height, parking ratio, energy efficiency status, real estate development firm, etc. — multi-tenant office buildings designed by top architecture firms do not fetch higher rents.

Results were encouraging, suggesting that buildings designed by top firms do exhibit a rent premium of 3.7 percent on average across my observed sample of 8,500 buildings.* Encouraged by these results, I prepared to approach the world with my findings. I could now correlate a statistically significant value premium to top quality architectural design!

Professor Redfearn cautioned me to dig a little deeper before making any ambitious pronouncements. One bedrock of statistical analysis is the difference between disproving a null hypothesis — which I had done successfully — and actually quantifying an effect. In that respect, this research was just the tip of the iceberg.

Upon closer scrutiny, the results began to raise as many questions as answers. While a strong premium exists for Class A office space in the Los Angeles and Washington, D.C., markets, the same premium is unstable in other markets we measured. Interestingly, I also found that this issue of unstable results is not limited to my own analysis, but those of the precedent studies which I used to build my model framework. Our finding of broad evidence for the impact of design on multi-tenant commercial office buildings, while not incorrect, is just a jumping-off point.

This makes sense. Modeling real estate is an extremely difficult — some would say impossible — task given the multitude of factors and decisions that impact asset and rent prices. Even after the numerous controls we put in place for our analysis, site-specific drivers of value are not fully captured — location factors such as amenity offerings, address prestige along with other factors like tenant mix, building efficiency, terms of existing leases, and the health of the local economy — resulting in a lot of uncertainty in the data.

Instability does not negate the existence of a premium. Given the complexity of the data, it has understandably proven difficult to isolate and quantify a “pure” impact of design. This is simply a reminder that statisticians, including my colleagues and myself, need to continue efforts to isolate and analyze the design factors that drive value. I continue to believe that the intelligent use of statistical analysis can bring us closer to finding those relationships.

This is where my continuing efforts into Big Data analysis are focused. Many of the value drivers of commercial real estate are correlated to each other, and many might contribute to rents in nonlinear ways. Borrowing the words of Nate Silver, Big Data will certainly deliver more noise than signal at first, but I am confident that the pure impact of design on value is within reach.

In his recent book, Silver outlines the pitfalls of being what he calls a “hedgehog,” someone who approaches empirical analysis with a predisposition to use the evidence to support what they want to believe. In contrast, “foxes” are those who are unrelenting in their search for meaningful results and are willing to accept refutations of their hypotheses. Architects, statisticians, and consumers need to remind themselves that while results might support a belief, it is critical to look below the surface to see if the devil is in the details.

While seeking fox-like tendencies, in certain ways I fervently remain a hedgehog in continued pursuit of what I know to be true: that top-quality architecture matters in our lives. The experience of working in an exquisitely designed building (take a look at where I work if you don’t believe me) enhances one’s life more than working in a cardboard box on the same site. We can never forget this. Balancing this is an openness to new relationships to understand how and where that value lies. Stay tuned.

*Results of the analysis suggest a mean rent premium of 3.7 percent with a standard error of 0.008, and a greater than 99 percent level of confidence; the 99 percent+ confidence interval is approximately 1.6 percent to 5.9 percent.

Chris Jerde has had a number of unique professional experiences, from working at the first Johnny Rockets hamburger store in Los Angeles to tending the music archive at the Smithsonian. But his real passion is how design affects everyday life, and since 2005 he has explored this question as a member of Gensler's Consulting practice. Contact him at chris_jerde@gensler.com.

Reader Comments (4)

Chris, thank you for this excellent post and your incredible research efforts! I look forward to learning more about what you find in the next steps of your research!
04.29.2013 | Unregistered CommenterErin Cubbison
Great blog Chris - I guess one of the burning questions is how can good design increase value by more than single digits?!
04.30.2013 | Unregistered CommenterPhilip Tidd
Hallelujah, brother and Amen! Absolutely agree it's the tip of the proverbial iceberg and accordingly, you're onto something really huge. The endeavor to isolate and quantify the impact of design is a challenging but worthwhile one and as you suggest, will provide hard evidence to support what we instinctively know to be true (that good design matters!) Rent premiums are but one metric for assessing the positive impact of design, and an excellent start to researching this topic... taken in aggregate, I strongly believe all the qualitative and quanitative benefits will build a strong case in the end. Exciting and inspiring... keep up the great work!
04.30.2013 | Unregistered CommenterJane Greenthal
Chris, thanks for sending signals amongst the noise. I guess I am a hedgehog as well, when it comes to the importance of design and top quality architecture in our lives, but the fox in me wants to understand more.
05.3.2013 | Unregistered CommenterDavid Willett

PostPost a New Comment

Enter your information below to add a new comment.
Author Email (optional):
Author URL (optional):
All HTML will be escaped. Hyperlinks will be created for URLs automatically.