Optimising the hyperlocal, and the macro, in real estate development


Real estate has long relied on instinct and retrospective data. But place-makers and decision-makers aren’t really thinking about the past, or even the present. Across the industry, long gestation periods mean that developers usually have their sights set firmly on the future — the long-term prospects of the property, of the district, and of the people who will live and work, co-live and co-work in neighbourhoods and cities that have the propensity to change overnight.

Forecasting district futures with confidence

In 1994, Singapore reclaimed the final 38 hectares of bay front land to create the Marina Bay we recognise today. Hong Kong’s own Starstreet Precinct has evolved into its current avatar over the last decade, and continues to change. In Bangkok, The Commons answers the challenge of designing a retail space sensitive to the changing needs of city dwellers in a tightly packed, tropical environment. 

District, street, or building, urban spaces are the outcome of years of planning and foresight. Foresight however, is messy business, especially given that two properties seemingly identical on the surface (when evaluated by traditional metrics), ultimately experience very different growth trajectories. The ability to tap lifestyle patterns, to map demand over longer time spans, and with granularity when it comes to predicting tenant expectations — these are the perspectives developers need.

Fortunately for us, data science is finally making a dent in real estate. We’re now equipped with not only the data but the right tools to better analyse cities. We’re getting better at answering the questions that real estate players have long grappled with: Where do we build? Who do we design for? What amenities do we offer?

The key is tapping into non-traditional data — hyper-local and subtle macro behaviours that drive nearly 60 per cent of the differential in real estate outcomes.

How do you map a place, and its people?

Successful urban developments rely on meaningfully understanding the decision drivers of current and future inhabitants. Because narratives of neighbourhoods already exist — in the minds of its people. If we could ask the questions, there are things we’d like to know.

Where do the locals eat? Where do they work? How do they spend their time after hours? Where do the tourists stay? The local boutique hotel or the chain? How important are outdoor spaces?

And simultaneously, we’d like to track how these patterns are shaping and reshaping the district, changing spaces and resource use.

What amenities on-property are rising in demand? What are the emerging expectations in tenant mix? Which outdoor and natural spaces are the most frequented? Which are the least used? Where does the district rank in terms of health & wellness? What is the district’s liveability score?
Capturing these nuances means that real estate planning can be targeted in more strategic ways.

When AI and human intelligence meet non-traditional data

Spectra FocusTM goes beyond the handful of data points developers usually have access to through government sources (lot sizes, prices, demographic data) and zeroes in on emotional and behavioural metrics that explain current and future space use. It does this by incorporating cues from a vast pool of sources — Facebook and Instagram for personal behaviours, gourmet review sites for F&B trends, home purchase discussion forums for residential considerations. Now we can ask questions, and the data points us to the answers. We can tell how many popular cafes a district has, how many well attended events, we can tell how resident expectations differ from those of visitors.

Leveraging many millions of relevant data points, FocusTM relies on a combination of Artificial Intelligence, Natural Language Processing and human analysis to help clients develop a strategic understanding of districts. The platform goes beyond a purely AI-based approach. Once AI has filtered conversations down to the 500 or 1000 most relevant samples, FocusTM relies on a team of analysts to manually sort and hand-tag conversations, identifying qualitative, location-specific nuances, trends and correlations based on their deep subject expertise. Neighbourhoods with identical quantitative and demographic metrics may in fact be poles apart when you look closer. Pairing AI-based research with human analysis, gives us the ability to answer far reaching questions, study competitors, even analyse comparable districts in other geographies to map tenant expectations and turnover.

With a platform like FocusTM, it’s possible to develop a visual map of the district based on consumer actions, gathering and making sense of activity, exploring the interrelationships between people and spaces — whether public, retail, outdoor, residential or commercial. 

Unearthing real insights before the opportunities are gone

“In conversation with our real estate clients, we’re often asked how granular we can go,” says Simon Chung, Insights Director at Spectra Partnership. “It’s possible with FocusTM to get a much closer read of the district landscape and the competition, to make informed decisions on everything from land acquisition and property development, to asset repositioning and disposition. We’re shrinking the set of unknowables, and giving developers an intuitive grasp of what they will have to contend with. The other important differentiator is speed — with traditional research, by the time investors can process data and distill action, the best opportunities have come and gone.”

FocusTM can identify gaps and opportunities in the context of the neighbourhood. The platform makes it easy to extend a study longitudinally to track how districts and people evolve. This has been particularly relevant after COVID-19 — district landscapes and sentiments are undergoing rapid changes and traditional approaches can’t possibly keep up with the pace of transformation.

“When we study a destination district like Sentosa in Singapore, we’ve looked across 24-plus months to see what brands, events, locations are moving up and down in terms of visitor interest,” explains Simon. “Or for a more community district like Discovery Bay in Hong Kong, we can identify how residents are interacting both in their own ‘home’ district and also what they do when they travel outside it. This gives us clear insights into the types of experiences and activities that could be brought into our district to better serve their needs and desires.”

By effectively aligning our built environments with the priorities of current and future inhabitants, big data is bringing real insights to real estate. We’re changing the future of property development as we know it.