Better design decisions: Architecture and data

Data is transforming business – including the creative industries. Daniel Davis explains how architects could be leveraging data to produce better design outcomes, and why some architects are making the switch to data science.

Data is the language of modern business. It is so fundamental to the 21st century economy that some have called it the new oil – a crude by-product of our digital lives that is streaming through our internet routers, sloshing about in our smartphones, and fuelling the growth of the world’s largest companies.

All of this data has one crucial purpose: decision-making. Which news story should be shown first on the Facebook feed? Which investment strategy yields the best returns? What will happen to hospital capacity during the pandemic? These are all decisions that benefit from the empirical evidence that data offers.

Which makes you wonder: if data is crucial to modern business, where does this leave architecture firms?

Most architects don’t have a natural fondness for data. They’re moved by sketches, not statistics. They’re comfortable making decisions intuitively and often look incredulously at the spreadsheets that developers present. This isn’t to say that architecture firms don’t deal in data. They do: modelling data, performance data, financial data … But the majority never bother to collect it, analyze it or incorporate it into their decision-making processes.

The pressure for architects to adopt data-led approaches is gradually increasing. Clients have seen analytics transform other aspects of their business. Developers are using data-driven tools like Archistar to more accurately assess the financial viability of sites. Real-estate agents are using data-driven marketing platforms to more effectively target potential owners and tenants. And operators are increasingly filling their buildings with sensors and IoT devices1 to better understand how their spaces are performing.

Some might say that architecture is different because it is more creative than development or real estate. But other creative industries, from filmmaking to advertising, have already been transformed by data. These days, most advertising dollars end up at Google, Facebook or Amazon,2 which aren’t creative, Mad-Men-esque advertising agencies but, rather, tech companies using data to match people with ads.

Architecture isn’t that different. Sure, there is a creative element to it, but there is a pragmatic component too. Surveying the built environment as it stands currently, you’d be hard-pressed to say that our present decision-making process, based mainly on heuristics and intuition, delivers the best results for people and the planet. So, it seems inevitable that architects will turn to data and analytics in an attempt to create better design outcomes.

Architecture firms that have attempted to become more data-driven have tended to run into two significant impediments: a lack of data and a lack of people to analyze it.

In a world teeming with data, it may seem surprising that anyone has a problem finding data to analyze. After all, architecture firms produce enormous quantities of data. For example, most drawings come from building information modelling (BIM) software, which is nothing more than a fancy interface for a spreadsheet. But unfortunately, these models are rarely consistent from one project to the next, making it hard to analyze the data they contain.

Nathan Miller has run into this problem a lot. He is the CEO and founder of Proving Ground, a US-based consultancy that helps architects and others in the building industry use data more deftly. Miller says that these projects typically begin with a conversation: “Our clients will come to us and say, ‘Hey, we’d really like to leverage data more in our design process.’ And we ask: ‘What data do you have?’”

Often, the answer is BIM data. But Miller says that while firms typically have a lot of BIM data, it’s often “not up to the point of quality where it makes for good analytical content.” To use the data, they have to clean it. “That becomes a hard pill to swallow because these firms have a lot of data, and much of it is a mess, and now they have to invest time, money, and energy to clean it.”

Of course, firms aren’t limited to analyzing BIM data. Depending on what they want to know, they may choose to study project performance data, geographic information systems (GIS) data, or even their internal financial data. But whatever the case, the general point still stands – most architecture firms aren’t sitting on mountains of data ready to be analyzed. Before they can even start the analysis, there is a bunch of housekeeping required to gather, clean and prepare the data.

For firms that manage to get their data into a usable format, a second challenge awaits them: they need to find someone to do the analysis. The popularity of data-driven analytics means that everyone from Google to the government is trying to hire people fluent in statistics, making the market for talent fiercely competitive.

Some architecture firms try to sidestep the competition by upskilling existing employees or outsourcing the analytics to another company. Neither is a perfect option. Upskilling has its limitations because statistics is a vast field that takes a long time to master; it’s not the sort of thing a designer can learn over a weekend. On the other hand, outsourcing comes with its own problems– namely, if you believe data analytics is the future, why would you give it away?

Some architecture firms are hiring data analysts directly, but with data science being one of the fastest growing professions,3 and the pool of candidates limited, competition and salaries are climbing. According to Indeed, the average data scientist in Australia earns $108,000 per year, compared to the average $83,000 made by an architect.4

One architect who made the leap to data science is Carlo Bailey. Having once worked at Foster and Partners, he’s now a senior data scientist at a real-estate analytics company. I asked him whether he’d ever go back to the architecture industry. His answer was direct: “I wouldn’t go back to a typical architecture firm.”

Bailey reeled off several problems with architecture firms – a lack of managerial support (“the amount of face-time I get with managers in other industries is way higher”), poor operational efficiency (“I haven’t filled out a timesheet since I left Fosters”), and lack of growth opportunities (“there is a glass ceiling [for people who are technically inclined]”). He also worried that his skills wouldn’t be put to good use because of the limited data and types of problems that architects solve. But he did allow that a non-traditional firm or a design-build firm would be interesting. “They can leverage data scientists because they have clearly defined metrics and their projects are repeatable.”

It’s not exactly a secret that architecture firms sometimes have brutal cultures, with long hours and relatively poor pay. Practices have been able to get away with this because architects don’t have many other workplace options. But a data scientist can work anywhere. If architecture firms are to attract and retain data analysts, their conditions need to stack up against those in other industries.

Given the difficulty of finding analysts, many of the leading firms are sourcing them in whatever fashion they can, whether that’s outsourcing, upskilling, hiring, or all of the above. “More than anything, we see a combination of all three,” says Miller.

In a lot of ways, this is a situation reminiscent of the early 2000s, when many firms were beginning to adopt BIM for the first time. Practices hired specialists and bought software, assuming that BIM was something they could just add to their existing processes. Many of these initial efforts failed because BIM required a new way of working. Firms spent years retraining staff, updating processes, and undergoing the cultural transformations necessary to make BIM successful.

Today, data analytics appears to be in a similar place. Some firms are hiring data specialists and buying analytic software in a leap of faith. But becoming a data-driven organization requires more than people and technology; it requi r es a cultural transformation. Firms need to begin actively gathering data, they need to address the cultural issues that make the industry unappealing to outsiders, and they need to be open to making decisions with information rather than intuition. For the firms that manage this transition successfully, data looks set to transform their business in much the same way it has reshaped other industries across the globe.

1. IoT, or Internet of Things, refers to any physical devices – such as smoke alarms, thermostats and fridges – that are connected to the internet.

2. Brian Wieser, “U.S. advertising market update,” Group M , 26 March 2021, groupm.com/united-states-advertising- market-update (accessed 20 September 2021).

3. Lisa Marie Segarra, “This is the fastest growing job in America. Here’s how much it pays,” Yahoo! Finance , 5 April 2018, finance.yahoo.com/news/fastest-growing- job-america-apos-165253226.html (accessed 21 September 2021).

4. See au.indeed.com (accessed 20 September 2021).

More discussion

See all
Bathing, ritual of healing and purification, is at risk of disappearing from our dwellings, through careless design. In praise of the bathtub

The humble bathtub is fast disappearing from our dwellings. Elizabeth Farrelly explores how the ritual of bathing is being erased through design.

The Greenary in Italy by Carlo Ratti Associati in collaboration with Italo Rota. The House of Green: Natural Homes and Biophilic Architecture

The House of Green presents the residential work of architectural practices that are treating the built and natural realms as one cohesive entity rather than …

Most read

Latest on site

LATEST PRODUCTS