In recent discussions with industry leaders and as outlined in our recent thought leadership piece “Evolution of Real Estate Network Planning in the Age of AI”, GeoAI is an emerging fast growing field of study enabling real estate and GIS professionals with next generation tools for market analysis and retail network planning. Outcomes such as sales forecasts, profitability estimates, ideal markets to locate in, foot traffic estimates, and a host of others that are predictions formed using complex neural networks and other AI type algorithms are now here to help increase the speed and accuracy of business case development and real estate scenario planning.
Technically, GeoAI is the integration of two well-known capabilities - Geospatial Data Analysis (or GIS) and Artificial Intelligence technologies. An editorial published in 2019 in the International Journal of Geographical Information Sciences by Janowicz et al. explains while GeoAI is emerging because of advances in computing power, vast data availability, and new analytical methodologies are all key ingredients, but the change in the “culture of sharing” is even more important.
“...there may be something even more important than the pure availability of data and advanced methods combined, namely a change in culture.” (Janowicz et al., 2019)
The culture change is all about data availability and share. It is defined by Janowicz et al. on three aspects of behaviour:
- Industry and AI practitioners are not as sensitive about protecting their data, in fact sharing it openly is becoming widely beneficial in a “...new data economy.”
- Reusing data is the new normal. This may seem like a trivial point from today’s perspective, but data reuse at scale is a new concept for many scientific domains.
- A new paradigm of data-intensive exploration highlights the increasing role of data synthesis alongside analysis.
“While none of these three identified aspects alone is necessarily new, the arising data culture certainly is.” (Janowiz et al. 2019)
Further to the methodological advances, an interview with Dr. Wendy Keyes Weniger, Principal Data Scientist: Spatial Science and Big Data Analytics Team at ESRI, highlights the point that GeoAI is the product of these paradigm shifts mentioned above. In this interview, Dr. Keyes Weniger defines GeoAI as:
“GeoAI is geospatial artificial intelligence, and it’s a kind of artificial intelligence or machine learning that’s used to simulate future outcomes. It runs on GIS technology [a geographic information system], and it often draws on statistical modeling, computer vision, and simulation tools.”
GeoAI gives us the capability to leverage rich data sets that include multiple dimensions; time series, cross-section, and now spatial data dimensions! The pure volume of data available now necessitates the use of AI and machine learning technologies that are fast and are capable of identifying important relationships in the data across all three dimensions. Data scientists and GIS professionals can now build models that are “geographically sensitive” where to be sure, the local market characteristics and behaviours can be dramatically different! For example, predicting mature new store sales and profitability now goes beyond aggregate time series projections, they are now sensitive to the unique dynamics in the local market - competitive density, traffic patterns, demographics and socioeconomics, etc.- and can be somewhat different by region and market.
In addition, we can integrate models calibrated on historical data with “generative AI” data about the future course of certain variables that are important in the models, but we cannot control directly such as, GDP, immigration, development policy, inflation, and a range of other macro variables that impact real estate decisions by retailers and developers. Leveraging GeoAI models, and future cast data about markets and economies (a.k.a. hypotheses) as inputs to the scenarios, enables “what-if” type simulations and business case scenario analysis. With a system like this in place, the ability to run a range of defined scenarios can begin to give executives realistic range estimates of its investment ROI - worst to best case - and inform debates at real estate approval meetings that focus on the risk-return tradeoff rather than whether one forecast is right or wrong!
The majority of enterprise executives acknowledge the use of scenario analysis, which involves discussing future outcomes — an exercise commonly practiced across organizations. However, this process often involves utilizing a multitude of basic tools, all of which feed into an extensive Excel spreadsheet. The absence of automation in this practice leads to increased risks of errors and unknown accuracy levels, posing significant barriers to its broader adoption and effectiveness. The result is that decisions tend to revert back to more “gut” instinct by those with the most experience.
We heard during our research with Executives that scenario planning using “GeoAI-type” simulations, as described above, for Real Estate planning has wide appeal - but most think it is too hard and again, trust will be hard to gain. Their lack of trust comes from two sources - missing data usually owned by brokers and the lack of talent for the AI/ML model building process. We dig into these topics in our Ebook, but if AI/ML models can get you 80% of the way there within minutes, this has real value. AI will never replace brokers, they have too much local knowledge about Real Estate planning and site selection that is not "in the data". They will always be necessary, but can be supported by an Intelligent system (a co-pilot) to get them 80% of the way there, enhancing their value, not hurting it.
Furthermore, the talent issue is real. The interview with Dr. Keyes Weniger highlights this challenge that we heard over and over in our interviews:
“....the lack of intersection <of data science and GIS> tends to be the problem. GIS professionals now have access to analytical tools well beyond the classic capabilities they trained on. And most data scientists haven’t been trained to understand spatial analysis. To make things more challenging, each group uses a different language, and even when their vocabulary overlaps, the words often mean different things.”
This is where the talent barrier mentioned earlier really comes into play. As GeoAI gains traction, GIS and Data Science teams will be forced to work together more directly and invest in technologies, cross-education within a business context to build and enable network scenario planning to inform strategy. We heard from executives who we called “AI High Performers” that have taken it to the next level by building a “Center of Excellence for GeoAI” as part of their AI strategic imperative to support growth and optimize brick and mortar networks.
The challenge is that most retail Real Estate organizations are not “AI High-Performers” and are still using traditional GIS mapping applications as well as MS Excel tools to run demographic reports to do site selection. This will change over time as the barriers to adoption of GeoAI capabilities come crashing down. A few educational institutions (e.g. University of Southern California) are already offering specialized Spatial AI programs demonstrating the need to build new skill sets for the modern world.
At PiinPoint, we have embraced GeoAI as the foundation of our software and go-to-market. We recognized early that providing an easy to use and intuitive spatial analysis platform that embeds GeoAI model systems for our clients was an unmet need. In fact, as a deliberate part of our implementations and delivery of custom modeling approaches, our Network Simulations engine allows clients to simultaneously string together time based real estate actions such as opening a new store, closing, renovating, relocating, consolidating two or more stores and quickly predict the net impact on top line and profitability of the network.
Check out PiinPoint. PiinPoint exists to support our clients' need for accurate and timely input to their real estate approvals discussions to increase their trust in their Real Estate market analysis, forecasting and network planning processes powered by GeoAI.
