I wrote about his issue at length for ESOMAR in 2013. The problem is that we’ve put the cart before the horse, IMHO. The data itself is just data and often has lots of problems, including the one noted. The belief that the data will actually tell you anything is misguided — you have to ask it questions. Lots of companies have invested heavily in kitchen sink approaches — let’s throw everything into an analysis and see what pops out. I advocate the reverse. Ask specific questions of the data and then make sure you have the right data to answer the question. A retailer might want to look at co-purchasing, either at the transaction or customer level, to see if there are synergies between businesses. Think peanut butter and jelly, or cake mix and frosting, both pairs being purchased by the same people or at the same time. If the former, promote at the same time. If the latter, promote only one of the pair at any one time (because they’ll buy the other anyways). Just… Read more »

This is a timely topic for me. With several bank and credit cards accounts (personal and business) at a large U.S. bank, I had a situation this morning where the bank told me I suddenly cannot use their online bill pay to pay a bill for one of their own credit cards, which I have been doing for five years. They say it could be that their system is screwing up, or it could be I need to use another browser or it could be … whatever. They are terribly sorry for the inconvenience but try again later or call back later … As happens with many retailers, this is a case of too many tech/data moving parts and interactions and too many customer service reps who are just programmed to give pat answers. On the other hand, I do find Amazon recommendations to be getting better (although perhaps that’s because I am giving them more information), but their customer service isn’t working as well as they have apparently offshored many of their call centers.… Read more »

Many retailers are making huge advancements by leveraging Big Data with the right tools and user interfaces that help decision-makers. Big Data must be coupled with the right decision-making tools — sometimes that might mean a simple-to-use front-end or user interface or an API into the retailer’s forecasting toolset or apps.

Big Data needs integration into the decision process. I am familiar with one major apparel retailer that focused their Big Data efforts just on inventory performance — real-time. They had a user interface for the merchants that let them see where the demand was and where the inventory was. This help them re-balance the next replenishment efforts and assured that items most in demand would be in stock.

Simple? Not across multiple countries and thousands of stores in real-time.

Big Data becomes big knowledge with the right platforms and tools.

Ah … George has repurposed that old Aristotelian argument about potentiality and actuality and somehow applied it to 21st century retailing. Bravo! As Aristotle told us, potentiality mist precede actuality, i.e., it must be possible for a thing to happen before it can, in fact, happen. So, in the case of “Big Data” its potential is clear even — as George correctly points out — if its potential is far from being realized. Too often we forget that individual datum, and therefore data, are both content and value neutral. Put another way: absent an effective contenting vehicle, information has no intrinsic value. And so, while we can take some measure of pride in our ability to aggregate some data, we are woefully lacking when it comes to exploiting the promise of “Big Data.” It isn’t capture that’s critical, it’s context. As to the future, if we keep confusing the ability to aggregate information with actual information, we aren’t going to make much progress. Hopefully, more people will begin to understand the paradox George has surfaced — the more… Read more »

The problem with Big Data is that people try to analyze it just as they did with “small data”. You can’t. It is too big. It has to be used very differently. You can’t look at the numbers uniquely. There are too many. One must find a different way to analyze the data. It has be done graphically.

There is a great book that uses Big Data graphically, The Human Face of Big Data. Also, there is a group in Sweden that is spectacular with its graphic interpretations of Big Data, http://www.gapminder.org. Try it out. It is so much fun to use.

Big Data is not going to tell big retail how to specifically sell to John Smith. Big Data is going to tell big retail where the business is and where it is going. Then the retailer has to make decisions.

This is the classic “the companies you’d expect to be leading innovation are the ones at the forefront with using Big Data and analytics” scenario. With its use of dunnhumby, Kroger is certainly leading the way with taking large amounts of data and converting it into actionable knowledge on tactics like targeted promotions and price optimization. Macy’s is deploying big analytics to not only engage with the customer better, but to improve the potential of new items. And there are many more examples. One of the other things I’ve recently heard a lot about is retailers using the data to avoid mistakes, either their own or others. Strategic and competition units within retailers are reviewing what’s gone wrong in the past and creating models to ensure the problems aren’t repeated, and even trying to reverse engineer them to see why they failed. I do understand George’s reservations about Big Data, though. More than once I’ve thought that even with all the data that, say, dunnhumby collects there might still be an analyst sitting in a… Read more »

George has identified two essential steps that Retailers have to take. That will entail stepping back from protecting the system that they have in place. In terms of insights, George’s perspective is spot on. Target, and too many other retailers have superb insights into what consumers do in their stores, but they are clueless as to what consumers are doing away from their stores. The data scientists that are being hired by retailers have to keep in mind the second portion of their titles — “scientists.” That will require them to explore and test more readily, as opposed to solely defending the collection methodology of data used in the past. With the wealth of multiple data streams available to data scientists, there is limited need to permit analytical inertia to creep into the mix. Analytics is not a black box service. By letting today’s computers cross and append diverse data sets — proprietary, custom, syndicated, government, industry — retailers will serve themselves well by uncovering the new “new,” and provide themselves with a path that only they… Read more »

It is not so much about Big Data as it is figuring out the right data. Have you read Malcolm Gladwell’s Blink? A great read about applying knowledge and wisdom to data. Then culling it down to the right data to support your actions.

But that’s just my 2 cents.

They merely need better collection, mining and utilization of highly personal Big Data? Oh please. This is the shiny object excuse too many companies now use to explain their own time-tested retail failures: not supplying adequate or respectful customer service, not having shelves/warehouses adequately stocked and having slow or aggravating checkout procedures.

I have worked with senior retailers every day for years, and never once have I heard one of these leaders say, “I need more data.” In fact, they are drowning in data. Merchants, eComm directors and marketers head into the dreaded “Monday reporting meeting” with their own spreadsheets and analytics/BI reports, disconnected from one another in every way. Rather than being ready to take the actions that will improve their customer experience and answer their customers’ true desires, they are into Wednesday or Thursday before they know what truly happened the week before. They are then left arguing their own theories of what action should be taken to solve for their individual KPIs, in their finite area of the business. Big data is only useful when it is actionable. When it is connected across channels and in every area of the customer’s shopping experience. When retail teams are looking at one “source of truth.” When they know on Monday morning where disconnects are happening between marketing and inventory, sales and returns, etc., and when they… Read more »

There is nothing more profound as a simple solution which has been properly prepared, implemented and maintained. In addition to redundancy and authentication problems there is, as the discussion implies, a matter of material relevance which is taken into the now and future decision making processes for marketing strategies. Is the data we are collecting coming from asking the right questions in the proper order? I think not, for if it were, the data we own would be seen as important instead of just big. This in turn drives us to collect mountains of data that is less reliable causing retailers to know almost nothing about 21st century shoppers. The opportunity is to know exactly what we need from an inquiry and what questions to ask and the order of asking to get this information. The first step is to identify all intuitive data/information and the unsupported assumptions that key executives carry with them into the company’s business decisions. The next step is to “learn how to learn” with the modern tools that are both… Read more »

To George’s first point, some retailers are grabbing some great insights via partnerships with other merchants of all types to share data via common loyalty platforms and other vehicles. Further, there are now more capabilities available to leverage online chatter in new, innovative ways that helps fill the gaps in knowledge of a particular retailer. Putting all of these channels together develops a much more complete picture of the shopper.

Point #2 is all about data being the new “oil,” however it is only as good as your tools are to derive actionable insights from the data. E.g., a barrel of crude oil is not very useful to run a car, unless you have the tools to refine it. The software in the marketplace today, along with the still critical human interaction element can grab those usable insights to make decisions based less upon anecdotal knowledge and more on science. That is the key.

It’s not the data. It’s the interpreters, and managers who don’t manage it on a granular basis. Big data is nothing. Teasing one-to-one insights out of it is everything.

If you look at where retail is going, and the data-driven capabilities required to operationally enable the retail enterprise of now and in the future, the continued convergence of web and stores, the connected customer—technology enabled to buy anytime and anywhere, Internet of Things for retail (starting to move along now past the hype stage, to in-store operational deployment by some Fortune top 10 Retailers), it’s not a big leap to conclude that Big Data and Analytics are here, and here to stay, in our wonderful world of retail. According to a recent survey fielded by Forbes Insights and Teradata, 78% percent of the 316 companies surveyed see effects of data-driven insights on revenues. Retailers, in particular, see an ever greater effect on their bottom line, with 31% of retailers seeing a revenue increase of more than 3% from data analytics insights, compared to 29% for other industries. Other retail industry findings from the survey: Data analytics is a key CEO focus for 37% of retail CEOs, versus 17% of CEOs in other industries Data and analytics… Read more »

I know of may retailers using so-called Big Data to improve store clustering, customer segmentation, assortments, allocation, promo/price optimization, etc. Where things get tricky is applying it down to the individual customer level, and although great strides have been made, it’s still immature.

I often get targeted for things I’ve purchased as one-off gifts, or items I’ve already purchased. The data set for an individual interacting at a specific retailer is actually quite small and therefore it difficult to derive meaningful intent.

We can already collect lots of general social data, so now we need to add purchase and behavior data from across all retailers. Then we’ll have enough data to form an accurate picture of individual customers. Maybe the next generation of payment solutions will help.

The issue from my point of view is how do you act on the data? Knowing and being able to respond is two different things. A lot of emphasis is made on getting insight from Big Data, but the ability to execute and deliver value to customer and retailer is still missing for the most part. To approach Big Data, the question should be “what is the problem I am trying to solve” and see if the data can guide actions, and then measure the results/accuracy. I see sometimes Big Data is “Let’s see what the data tells us and then we figure out what to do with it” which is the wrong approach.

If retailers truly want to extract value from Big Data they had best begin by abandoning everything they think they know about data accumulation, analysis and activation. The shopper data generated within the confines of the store are certainly very large, but the shopper data generated outside the store are far greater in magnitude and significance. It’s time to stop contemplating our (data) navels folks. We need to lift our heads and start gazing into the (data) clouds, if we want true insights into the Big Picture. Big Data is certainly about volume and variety, but most importantly it is about velocity. To cope with ever-increasing speed, we’ll need to stop attempting to measure the molecules in the data flow and start monitoring its temperature. In this context it is very often not the data itself that yields the insights, but the data about the data. All metaphors aside, I am at my core, a stores guy. I believe strongly that retailers must learn to offer relevant experiences and solutions to shoppers. Sensing their in-store… Read more »

I think there are core analytic/scientific problems with the theories of Big Data. Observed data (or found data, which is what Big Data is in this context) has exceptional weaknesses. In many ways, big data enthusiasts have gotten away with ignoring those weaknesses with the idea that “but there’s so much data, how can it not reveal huge insight?” And yet, that’s a fallacy. We’ve known for decades that more data doesn’t mean better action. It just means more data. But the core challenge in the fact that it’s observed. And observations never reveal motivation. Here’s a simple (not retail) example. A friend of mine worked for an accounting software company. They hired a Big Data firm to do all their prospecting. The firm analyzed the observable data and sent them the most active web folks around their product as valuable prospects. Except, the motivation for all that web action was that these were unemployed accounting people trying to learn about the software in order to get a job—they weren’t even close to prospects. Shift… Read more »

Hmmm quick quiz. If you could choose only one of:

a) A quant and a seriously awesome collection of all kinds of great data (in all sizes and shapes)

or:

b) A great buyer who understands the customer, fashion trends, and how to build a great collection for sale.

Which one would you put your money behind?

The great ship Panacea sailed some time ago. The correct answer IMHO is c) a well informed buyer… data and context.

Happy Friday.

Big data is like the Goldilocks story: you don’t want too much or too little, but just the right amount. Once you have that right amount it’s a matter of using the right tools and devoting the right team to analyze it properly. Next, a retailer needs to have a flexible enough strategy to implement the takeaways they gained.

I think that the potential of big data is vast, but it’s about finding the Golidlock zone for each individual retailer.

Currently retailers are doing their best to make the most out of big data, but there’s only so much they know about their customers. As they gather more and figure out how to analyze it more effectively, I see big data having a huge impact on improving the retail experience for shoppers and retailers.

Big Data spots trends. It doesn’t create a personalized experience. Something I refer to as “little data” or “micro data” drills down to the individual customer. That is where a customized experience—in your store and in your marketing—begins.

I see that many retailers are beginning to understand the importance of using more data to make better decisions. The first hurdle seems to be recognized by many, however, there is chaos across the industry on just how big data should be used. It’s been my experience that retailers do not need more data or more insights, in truth, I can not recall a retailer ever telling me they needed more of either. Retailers are looking for insight driven recommendations. If a retailer believes the net result of big data are “better insights,” I believe those who do will struggle to find measurable value in what they have invested big money to do with big data. However, if retailers are driving to recommendations that are delivered by technologies that leverage big data combined with advanced prescriptive analytics…now this is a winning equation. When the data of an entire enterprise is used and potentially combined with other sources of data outside of an organization (like syndicated assortment/sales data) recommendations can then include the impact to all… Read more »

The advancement—and opportunity for advancement—retailers are making by leveraging “big data” will prove to be some of the greatest benefits to consumers, since the ability to purchase goods/services from a connected device. There are few other plays in retail (at the moment), which lead to such massively beneficial impacts as collecting and leveraging “big data.”

This is evidenced and there is a very little debate, that “big data” is a macro driver of change to the consumers’ benefit in many industries already (transportation, healthcare, telecom, etc.).

Retail will absolutely rise to the opportunity to realize the benefits of “big data” for consumers, if for no other reason than because it is so commonplace (and beneficial) in nearly every other transaction in their lives that they’ll demand it.