Consultants Forum Advertise
Subscribe
Close

  • Free for qualified executives and consultants to industry

  • Receive quarterly issues of Area Development Magazine and special market report and directory issues

Subscribe Now
Renew

Area Development - Site Selection, Facility Planning and Workforce Development

The Promise of Predictive Analytics in Site Selection

Data science that encompasses analytically rigorous predictive models has the ability to disrupt the conventional location decision process.

Q3 2018
Data science is unlocking new opportunities for creating value across business enterprises. Leaders have always grappled with uncertainty as they plan and execute business plans — particularly when it comes to making critical decisions around relocating or expanding locations — but evolving data sources, analytical tools, and computational capabilities can now enable them to reduce that uncertainty and make better decisions than ever before. Learning from how data science is being applied to other business sectors, commercial real estate decision-makers can find new ways to optimize outcomes and create unexpected value for their businesses.

Advanced applications of new data sets have already shaken up the ways companies market their products and services. By combining deterministic customer data — which represents discrete verifiable information about a customer’s stated preferences — and probabilistic customer data — which uses algorithms to predict customer behavior based on similar patterns of behavior across populations and multiple devices — marketers can often suggest compelling offerings to customers before they even know they may want them. The results can be stunning. Coupons for baby food and diapers arrive before pregnancies are announced. Furniture ads pop up before homes are purchased.

In the field of commercial real estate, similar techniques are enabling some investors to better predict value swings in buildings and neighborhoods — producing more useful estimates of value than conventional hedonic analyses. Up to this point, most advances in the application of data science in real estate have been made by investors, while most site selection processes remain reliant on aggregating simple layers of deterministic data paired with more qualitative information on location reputation and branding.

Data That Can Predict Cultural Fit
Conventional site selection models have historically been driven by deterministic information related to verifiable factors like cost and availability of labor and real estate, transportation, quality of life, tax rates, and incentives. Data from these factors are collected, weighted, and analyzed to narrow down site alternatives and target locations that represent the best combination of a company’s preferences. Although most site selection consultants now use geographic information systems to produce beautifully layered heat maps and dashboards for visually guiding decision-makers, they rarely apply data science to the underlying variables to truly optimize solutions. And the static deterministic factors themselves provide limited insight into what has become the holy grail of many contemporary site selection pursuits — cultural fit.

New sources of deterministic and probabilistic information promise to substantially change the game for site selection. New sources of deterministic and probabilistic information promise to substantially change the game for site selection. Just as businesses are combining information derived from modern data aggregators like social media sites, peer-to-peer exchanges, wireless providers, and payment companies to drive marketing algorithms, real estate investors and appraisers are beginning to integrate information from these sources to identify previously indiscernible micro-market factors at scale.

Until recently, it has been extremely difficult to apply meaningful data science around lifestyle trends and unofficial hubs that may impact targeted recruitment and retention. But now real-time, aggregated knowledge about openings and popularity of hotels, residential developments, restaurants and bars, coffee shops, art galleries, cultural events, and festivals may provide unique insights into changing market trends and the potential flow of talent into specific areas. Similar knowledge about people’s movements, characteristics, and activities can provide rich measurements to help define the current and future state of an area’s labor, employment, culture, education, entertainment, transportation, pollution, and crime.

Stress-Tested Site Optimization Models
The transformative potential of such predictive analytics on real estate investment and development is enormous, but these advancements should prove equally profound in the site selection space, enabling users to predict the impact of important decision factors that may change over time and algorithmically optimize location decisions.

What if a company could base its location decisions not merely on an area’s historic information, but on its probable future characteristics as well? What if current widely used deterministic data could be integrated with probabilistic information from modern data aggregators to drive algorithmic site optimization models that can be stress tested to assess both opportunity and risk?

A shift toward analytically rigorous predictive models promises to enable trend-beating outcomes that can substantially surpass the value expected from conventional site selection processes. Just as data science has transformed marketing strategy and results, so too can it transform real estate strategy and the outcomes achieved by truly optimized site solutions.

Exclusive Research