WeWork's hot desk area in Seattle's South Lake Union neighborhood. (This is not the beta floor the article discusses, but it looks relatively similar.)
WeWork WeWork's hot desk area in Seattle's South Lake Union neighborhood. (This is not the beta floor the article discusses, but it looks relatively similar.)

By the numbers, WeWork’s dominance as a collaborative-workspace provider is hard to dispute. With a $16 billion valuation as of March, and more than 900 employees as of June, the company currently leases coworking spaces to 60,000 members at 110 sites in 30 cities across 12 countries. (It also has two WeLive locations, a residential take on the sharing experience.) This includes the 13 sites in seven cities that WeWork opened since the beginning of this month alone, which translate to 9,500 additional desks, each of which can be rented per month.

Aside from bringing in revenue, these hundreds of sites and thousands of desks also serve as one enormous test bed for WeWork’s 13-person product-research team to conduct massive studies in architectural planning, programming, and design. Kicking off this endeavor was the debut of the company’s flagship “beta floor,” the sixth floor of WeWork’s Times Square office, in New York, which hosted the company’s inaugural product roundtable last week.

“Physical space is our product,” said WeWork chief development officer David Fano, who donned a T-shirt emblazoned with the words “bldgs = data” for the roundtable. This mantra would be both the event's theme and a nod to the brain trust that WeWork inherited with its 2015 acquisition of architectural and technology consultancy Case, which coined the analogy.

At first glance, the beta floor looks like a typical WeWork space: industrial-chic finishes, wood floors, globe pendants, Sputnik-like chandeliers, and minimalist stools surrounding a bar stocked with coffee and fruit-infused water.

WeWork's dedicated desk spaces at its location in the Meatpacking District, New York (also not the beta floor in question, but similar in look).
WeWork WeWork's dedicated desk spaces at its location in the Meatpacking District, New York (also not the beta floor in question, but similar in look).


Then Alan Jackson, WeWork's director of systems research, began pointing out the space's distinctions. A compact, wall-mounted green box is one of 10 environmental sensors tracking the interior temperature, humidity, air quality, barometric pressure, and ambient light levels. White, wall-mounted beacons, 20 in all, distributed across common spaces (open desk areas and meeting rooms) triangulate the indoor position of WeWork members (the data is anonymized). And an overhead sensor, one of four, uses computer vision to observe member occupancy.

Last year, I wrote about the potential of the Internet of Places (IoP)—the use of the Internet of Things (IoT) and sensor technology to discern the success of architectural design—and described an in-house experiment spearheaded by Case to track its office's space use. Though Case is technically defunct, this research objective is being continued by WeWork’s product team, which comprises several former Case employees. IoP takes IoT beyond its current uses in smart thermostats, alarms, and lights to address a larger question that far too few architects ask: Does their design work for their clients?

Many factors preclude design firms from conducting post-occupancy evaluations: they no longer have access to their projects or their occupants; they lack a business case for studying what is likely a one-off design; and they, as well as their clients, have moved on to other commitments.

The situation is different for WeWork. The company has access to every square foot it has renovated and leased, as well as to the space's occupants. And it is important for the company to know how well a location’s “unit mix”—what it calls the ratio of private offices, meeting spaces, and open desks—works so it can tweak the formula at the next location.

One meeting room covered in a wallpaper with Rorschach-blob patterns, which some may consider as high design, was perceived as “distracting” and “weird” by some members.
Courtesy WeWork One meeting room covered in a wallpaper with Rorschach-blob patterns, which some may consider as high design, was perceived as “distracting” and “weird” by some members.


As Joshua Emig, head of product research, put it: WeWork aims for its product to “improve their members’ success, happiness, and well-being—and we want to prove it.” Though evidence-based design, commonly used in healthcare architecture, has a similar objective of using data and research from built projects to inform future design, Emig said the opportunity WeWork has to study its properties does not “exist anywhere else, on this scale.”

And, as WeWork's director of cities and spaces research (and ARCHITECT contributing writer) Daniel Davis noted, because each of the 110 sites adhere to the company's look and brand, the product team has some semblance of a control in the experiment of architectural design. That is, if each site uses the same furniture and coffee bar, for example, WeWork can focus on studying other variables, such as site, views, pricing, that make one location more popular than another.

WeWork also solicits feedback from its members on everything from supplies to building comfort through its member app. Comments are emailed to WeWork personnel daily.

Member feedback is sent to WeWork's headquarters on a daily basis.
Courtesy WeWork Member feedback is sent to WeWork's headquarters on a daily basis.


Space is at a premium for WeWork: With each new site, the company is gambling that its investment to overhaul an existing floor or building will pay off in new members. Determining a location’s particular unit mix is an art form that can have a real impact on its bottom line. Oversized meeting rooms and phone booths that sit empty could be rentable office or desk space.

Occupancy information is one low-hanging fruit for the beta floor to automate. A study published by Davis in March, which found that 12-person conference rooms are commonly occupied by two- or three-person meetings, was based on data collected firsthand by repeatedly walking by each room and manually counting heads.

Phone booths, available on a first come, first served basis, give WeWork members privacy.
Courtesy WeWork Phone booths, available on a first come, first served basis, give WeWork members privacy.
The utilization rate of phone booths is one variable that can be tracked through occupancy sensors.
Courtesy WeWork The utilization rate of phone booths is one variable that can be tracked through occupancy sensors.

Jackson showed one early result based on data gathered from overhead cameras on the beta floor: a heat map of the "hot desk" area, compiled using a week’s worth of images showing which seats people claimed the most (spoiler alert: next to the windows). He predicted that this computer-vision capability could someday be used as virtual turnstiles or to map foot-traffic patterns and optimize wayfinding. (After the roundtable, Davis noted that sensors with camera or computer-vision capabilities would never be placed inside private offices or meeting rooms to be mindful of personal privacy and intellectual-property.)

As technology becomes increasingly personalized and integral to everyday tasks, everything from mobile devices to electronic keycards can be used to pinpoint one’s location. Privacy, in many cases, has been traded for convenience. And, as Fano said during the roundtable, the WeWork members on the beta floor were unfazed when they were told that they would be test subjects in an extensive user-experience study.

WeWork's dashboard showing the occupancy and environmental data collected by IoT sensors in the beta floor.
Courtesy WeWork WeWork's dashboard showing the occupancy and environmental data collected by IoT sensors in the beta floor.


WeWork’s product team is just at the start of collecting and evaluating data. The environmental sensor, for example, doesn’t yet communicate with the beta floor’s HVAC system. As WeWork incorporates more technologies into their properties, Fano said he envisions someday saying to prospective members, “OK, this building is version 3.0 or 4.0 of a WeWork building.”

If WeWork does achieve its goal of connecting every location and every desk through IoT technology—effectively creating the IoP—what could this mean for the greater architectural community? Though Fano says the company is not opposed to sharing its post-occupancy findings in the future, Davis notes that the results would still be particular to WeWork projects. That is, WeWork’s collective portfolio—its design brand and fit-out style—is itself a one-off architectural case study with limited usefulness to designs by other architects.

Instead, Davis says, what would likely be most beneficial for the design community is WeWork’s methodology for collecting post-occupancy data through IoT. For example, what sensors are needed and where, how do firms gain buy-in from occupants, and how do firms collect and then parse the data? If so, WeWork’s greatest contribution to architecture may not be in spatial design or its giant renovation footprint, but instead in research process and knowledge.

WeWork's scale means it also has the ability to influence product manufacturing. The company uses miles of aluminum storefront systems to delineate spaces. WeWork director of building research Jason Andersen said the company has designed a frame whose section requires 16 percent less aluminum than the current product—saving both shipping and material costs. The company is also experimenting with the thickness of the glass lites to modulate reverberation.
WeWork WeWork's scale means it also has the ability to influence product manufacturing. The company uses miles of aluminum storefront systems to delineate spaces. WeWork director of building research Jason Andersen said the company has designed a frame whose section requires 16 percent less aluminum than the current product—saving both shipping and material costs. The company is also experimenting with the thickness of the glass lites to modulate reverberation.