Wine Data Visualization: Michelin Restaurants and High-Priced Wines

Screen Shot 2018-03-15 at 7.53.15 PM.png

Here's a peek at something we've been working on around the office lately.

(I write that and it sounds off-handed but, believe me, there's nothing flip about the effort that's gone into this process.)

The map above responds to a client request to understand consumer behavior around their wines in Michelin-reviewed restaurants in a target market, which in this case is Washington DC.

That was the question. To work toward an answer, our team took a few carefully considered steps.

We started with a data set of street addresses of Michelin-reviewed restaurants in Washington, and we mapped those. They're represented by the balloons.

That's the first data set.

Then we overlayed that geo-tagged information with a second data set, namely consumer scans of wine labels that are also geo-tagged. They're represented by the pins.

Then we correlated the latitudes and longitudes of those data records, which are accurate to within 50 feet, with the locations of the Michelin-reviewed restaurants.

Then we segmented by red wine, since that's the concern of the client.

Then we segmented by price point, since that's also the concern of the client.

Then we segmented by region of origin of the wine, such as the U.S. versus Italy versus Spain versus France versus Argentina.

Now we're diving into additional variables. One example is the competitive set, that is, other labels that were scanned within the limited parameters of the same session.

Another example we're diving into is consumer reviews of these wines, which our team can also map using our proprietary algorithm for hundreds of frequently used wine descriptive terms in the English language.

And so on, and so on, depending on the needs and questions of the client.

The map you see above is one small slice of this specific project, but the implications are huge when you consider how these processes can be extrapolated -- for wine retailers or grocery stores, for example, instead of Michelin-reviewed restaurants.

And so on, and so on, depending on the needs and questions of the client.

This is just one part of one project. There's a world of data out there, and we'd love to put it to work for you.

How can we help?

Please let me know your thoughts and questions, and thank you as always for reading.