Last week I attended the SAS Analyst event, one they hold every year in Steamboat Springs, CO. Among the four analysts at RSR, I’m pretty much a shoe-in for this event, given where I’m based, but this year I had more of an agenda than usual, as partner Steve Rowen and I are conducting RSR’s Business Intelligence benchmark report this year.

So, yeah, I’ve been hearing a lot of chatter about “big data ” and strange words like “Hadoop ” and I confess I haven’t been paying much attention. I usually like to operate at either an IT strategy level, or tactically, at the intersection between technology and process. I definitely try to avoid the deeper levels of the technology stack as much as I can – doesn’t mean that I can’t go there, but it doesn’t mean I like it there either.

But sometimes there are things that happen deeper in the technology stack that have strategic implications, and it seems that advances in technology’s ability to manage Big Data – thanks to things like in-memory analytics and game-changers like Hadoop – qualify as exactly that kind of situation: a hardcore tech thing that has big strategic implications. I have to give props to Brian Kilcourse on this one – he’s been nattering away over here both privately and publicly that this is a big deal. It took some sit-down time with SAS to get me to pay attention.

SAS opened its conference with an image that has stuck with me. They put up a slide that had a big lock on it, labeled Big Data, and a key, labeled Big Analytics. Now, in retail, I think SAS has been challenged to overcome a dichotomous reputation – on the one hand, they are well-known for their Merchandising solution, which is a very robust application – a real app, not a toolset or anything. On the other hand, they are known as the beloved toolset of the deep-dark analyst types that live within marketing analytics, teasing precious insights out of jumbled consumer data.

This idea of Big Analytics poses an opportunity to dash each of these singly-unfair reputations to pieces. If ever a moment was made for SAS, it seems like now is the time – the technology is there, the solutions are there, and the retail need is growing by the terabyte.

The challenge is getting the business side of the retail house to understand exactly what’s going on in this murky world of the tech stack and how it could change their lives. The reality of some of these improvements that are coming – optimize prices in 20 minutes instead of 5 hours, develop a demand forecast across every store and SKU in 90 minutes instead of 19 hours – is amazing, and yet these examples just don’t seem to do the opportunity justice.

I look at it this way. Right now, most retailers plan a relatively small number of key items and do so across an aggregated view of stores – A stores, B stores, C stores – maybe up to 5-10 different classifications of stores. The rest of the assortment is spread on a more aggregated basis, because no one in their right mind can truly plan every SKU across every store. Hold on to this thought for a moment: there is aggregation on both sides of the planning process – on the product side, and on the location side.

What if, with the same amount of people and in the same amount of time, you could eliminate that aggregation? Plan every SKU down to differences by department by location AND by customer segment? Price optimization and size/pack optimization have already revealed that a lot of value is hidden in the averages created by looking at an aggregated view of products. Assortment promises to reveal more – as does being able to look at each location uniquely, and not aggregated based on some model store. Now, among the merchandising providers out there, who legitimately has both the ability to disaggregate planning on the product side AND on the location/customer side? One company comes to mind right away: SAS. Between Merchandising and Customer Intelligence, SAS has the solutions that promise to be centerpieces in the fight over granular merchandise planning.

To me, this is the promise of Big Analytics – “high performance analytics ” in SAS’s parlance. When applied to merchandise planning, think of this – it took over 10 years before anyone started to feel like they had truly milked all of the margin opportunity out of a less aggregated method of setting prices. And in that time, the margin benefits they achieved were big enough that Wall Street sat up and took notice – started actually talking about the technology that retailers were using and not just the products they were selling or the stores they were opening.

If the industry can achieve the equivalent in assortment optimization, I think we’re talking about a bigger opportunity.

Hopefully it won’t take 10 years to get there.