The ace up the sleeve of Enolytics is, hands down, our data team. I sit in the room with them and watch them work, observe them interact with each other, and listen to them communicate in "their language" which is, still, new to me.

They are the data scientists while I am, for my part, a communicator. Which means that I'm interested in this language they're speaking, and in how I can go out and translate these terms from the world of data to the world of wine. 

Here are eight words and concepts about big data that I've learned since Enolytics began, that are helpful for making the leap of communication from data scientist to end user in the wine business.

I've organized the list roughly into the workflow that I observe happening in our projects. The scientists would no doubt describe it differently but, over the course of learning how to talk with wine people about big data, these are the terms and the sequence that make sense.

1. Query. A query is the request we make of our data partners that enables them to “pull” the records we need. Identifying the right query – asking the right question – often seems to me like the trickiest part of the process.
2. Anonymized. The raw data records we receive have been anonymized, or stripped of their identifying tags that would enable someone to trace a record to a specific user. This protects the privacy of the user.
3. Interactive Dashboard. We build interactive dashboards in order to transform the raw data from individual records into a navigable interface. The dashboard looks in many ways like a website, with point-and-click functionality.
4. Fields. The dashboard is comprised of areas of interest, or fields, that are significant to the project at hand. Depending on the project’s objective, a field could be anything from vintage to sales channel, and from brand name to zip code.
5. Package and Visualize. When we take raw data and use it to build interactive dashboards, we have “packaged” the data so that it can be accessed and seen, or “visualized,” in a more user-friendly way than the spreadsheet format of thousands of rows.
6. Aggregate. Layering one data source on top of another, such as U.S. census data on top of the zip code dispersion of a winery’s DTC customers.
7. Slice and Dice. Let’s say the client wants to understand the geography of consumer interest on a month-by-month basis. From the zip code field, they can select their top-performing neighborhood in, say, Chicago. From the calendar field, they can also select, say, October and/or November and/or December. Selecting and de-selecting portions or slivers of the data set allows the user to “slice and dice” according to their objectives.
8. Insights. Insights are, essentially, lessons learned. I see them as lightbulb moments that occur when information is presented in a clear manner that is newly available.

I hope this makes sense, and thank you, as always, for reading --

PS I was honored to be interviewed last week for this HarvestSummit post, where I spoke about things like the characteristics of Enolytics’ early adopters, the ROI of big data, and adding some wabi-sabi to our best-laid plans.