The Smallness of Big Data

This week we’ve been working on projects that involve big questions.

The overall perception of a certain category of wine in ten major markets, for example.

The launch of a new product within an already crowded category.

The effectiveness of a three-month marketing campaign in a particular market.

And so on.

They are big questions for big data.

But here’s the thing about big data: it’s small.

As in, millions upon millions of individual, unique records. It’s the aggregation of those records that makes the data “big.”

But in their individual smallness, the data enables us to get very granular and very specific about the insights into consumer behavior around wine that we’re looking to identify.

So it becomes a dance. “Small” records leading to big data, and “big” questions informed by small behaviors.

Let’s say we’re working on the overall perception of rosé wine in ten major markets, and the client wants to locate “hotspot” areas of interest. We consider the geography of where the data record was created; it has a latitude and longitude that’s accurate, in the case of mobile activity, to within 20 meters.

It’s one record, located in one very specific place. But when we look at individual records from the same very specific place, and each record documents a unique behavior, then we can start noticing patterns.

The inverse is also true. When we look at the same behavior occurring in different geographies, that too creates a recognizable pattern.

What big questions are you wrangling with right now? And how many millions of wine consumer data records can help you answer them?

I’d love to hear about it. Send me an email ( or give me a call (+1.702.528.3717) anytime.

Thank you, as always, for reading.