GEO-AI is arguably one of our most powerful solutions. Integrated in TargomoLOOP, it helps you find out how well a potential location could perform.
GEO-AI is the name of the software that builds a statistical model predicting the future revenue (or other KPIs) of any location, based on past data.
How is a GEO-AI model built?
Simply put, GEO-AI is fed with past performance data from our clients’ stores and we ask it to find meaningful links between this data and thousands of other data that we have curated.
What is this data?
- The locations of your competitors
- Statistical data that is either public or privately acquired
- Foot and vehicle traffic
- Points of interests
Any human with a bit of patience and advanced maths skills could do GEO-AI’s job. However, this human would probably run out of time and be very inefficient. GEO-AI on the other hand, runs hundreds of thousands of iterations on its own and is able to learn gradually what are the factors that make a location successful.
How do we know GEO-AI works?
You probably wondered: if the AI knows what the result for existing locations is, how can you actually trust its results? Even if you open a store tomorrow, it’ll take a year before you can see whether the GEO-AI predictions materialize, right?
Yes. However, when the GEO-AI model is built, we always “hold back” some of your locations. Out of your 200 locations, we will keep the revenue for 20 of them secret. Once the model is finished, we will compare the prediction that the model projects for these 10 locations with the reality.
The result? GEO-AI has shown repeatedly being able to predict the hidden locations’ revenue with anywhere from 85 to 95% accuracy. And the results keep improving as we learn to improve GEO-AI as well.
And should a client be dissatisfied with the output from a model or the statistics it uses, then we take the time to listen and adapt. GEO-AI is not a machine running on its own, but rather a guide with whom a conversation needs to be had to make the best results.
What is the output of GEO-AI and how do I use it?
The output of GEO-AI is a geo-spatial model that we integrate in TargomoLOOP. Every GEO-AI model is obviously presented and validated by our clients. Iterations are also possible.
Once the GEO-AI model is in LOOP, using it is very intuitive. It is meant to work alongside the existing features.
For any of your existing or potential locations, you can type in an address or find it on the map. Once you create the location, GEO-AI will calculate the potential revenue (or other KPI) for this location, taking into account the locations around. You can find the predicted revenue under the “GEO-AI prediction” tab.
The locations around it will also be impacted by this new location. You can check the impact on their revenue as it adapts dynamically.
Because we believe in transparency, the GEO-AI model will also show you all the variables that it uses and the extent to which each of them is used in the model.
The pie chart shows the influence of each factor on the final revenue.
And there’s more.
- If you want a full demonstration of GEO-AI, our founder Henning has got you covered here
- If you don’t know where to even start searching for a new location, GEO-AI creates heatmaps that not only tell you where a new store would be profitable on its own, but also where it would be profitable for your brand. Check this article (To be written)
- In the near future, you will be able to see where the revenue from GEO-AI comes from geographically. Which means: a more transparent way of taking decisions and better Marketing decisions, as you will be able to nail the postcodes and cities where you can get the most ROI from. Stay tuned and get in touch at email@example.com if you want to know more.