Originally posted by SiteZeus on December 11, 2018
We all hear about how A.I. is changing the world of retail real estate, and it’s true, companies like SiteZeus are using machine learning to predict sales revenue blazingly fast and accurate. However, intuition hasn’t lost its place. Real Estate directors know that sometimes the “vibe” of the location just doesn’t fit the brand.
The missing social criteria
But what does that even mean…“The vibe of a place?” Every location has its own personality and social fiber, but you won’t find “personality fit” as a site selection criteria. Before today there was no data source to measure the personality or social experience of a location.
In 2016, the team at Spatial.ai discovered the answer had been here all along. Every city emits billions of social signals – and no one was harnessing the immense but invisible layer of data that comes out every day. Leveraging machine learning, Spatial clusters every geotagged conversation from social media to create a Geosocial Taxonomy of the types of people that visit an area.
For the first time, site selectors can see beyond demographics into the social DNA of a block group. Are the people here interested in wellness? Are they trendy? How does this block group rank nationally for nightlife, artistic behavior, outdoors activities? Anything you can imagine people talking about on social media, it’s in the Spatial.ai taxonomy – ranked nationally for every block group.
What does that mean to leverage Spatial.ai data?
It means you have an advantage over the competition. When you use Geosocial data in SiteZeus you can:
- Reduce error in your predictive model
- Immediately light up the areas that fit your brand’s social profile
- Discover new positive correlation to new customers
- Select sites in target market more efficiently
How does it work?
Imagine you are a discount shoe retailer with hundreds of stores using SiteZeus. Two particular stores have stumped you recently. One store has record sales, and the other is far below Average Unit Volume and needs to be shut down. The confusing part here: both stores are essentially the exact same across your site criteria. Something else is going on.
Running Social Psychographics data through your predictive model surfaces some interesting results. You find the over-performer has a high index of “active moms,” a geosocial category that has a positive correlation with your brand. You then find the under-performer has a high index of “urban fashion,” a geosocial category with a negative correlation to your brand.
The under-performer was doomed from the beginning because it was not a social match to that community.
Perhaps the most useful thing about social network segmentation data is the story it tells. The insights in the data would take months of interviewing users to discover. In SiteZeus’ predictive models this richness is surfaced in seconds.
The over-performer was a social fit with the high index of “active moms” in the area – a segment that also tends to be “deal-seeking.” The under-performer is a social mismatch with the high index of “urban fashion,” a segment that is in a similar income bracket to “active moms,” but will spend their money on Jordans before a discount shoe because of the symbolic cultural value.
Spatial has worked with data scientists in companies like Ford, cities like Miami, Pittsburgh, and Grand Rapids, and retailers like Payless. Companies without sophisticated data science teams have struggled to truly leverage this data set. SiteZeus unlocks a whole new world of opportunity for brands. The data can be accessed immediately – and all of a sudden, the same insight that was historically only available to fortune 500 companies and cities is now available to any retailer using the SiteZeus platform. No messy data integrations or slow regression models. You get deep customer insights in seconds – backed by data. A new, unprecedented advantage for brands.