The National Retail Federation’s annual “Big Show” in New York is always quite an experience and this year was no exception. There was, of course, the Store of the Future, but a little more “off the beaten trail” I talked with a few technology providers. I wanted to touch on a couple of those ideas here, which highlight and validate the recent thinking we started to describe in last month’s “Fourth Dimension” and “Data Detectives” articles.
Know Your Geography—Know Your Customer
Retail, perhaps even more than other industries, lives or dies on differentiation, targeted at specific customer segments. Wal-Mart will never conquer the world (although they may become their own nation!), because there are just too many different types of people.
Understanding geographic segmentation is a critical dimension for successful differentiation. Our headquarters are in Harvard Square, in Cambridge, MA. Within just a couple of miles are vastly different neighborhoods in Back Bay, Somerville, Downtown Boston (which can be further subdivided), the South End, Allston, the Armenian district of Watertown, and the Italian North End, to name just a few. Each of these is so distinct and different—any retailer who hopes to succeed needs to understand the demographics and psychographics; population densities, attitudes, income levels, value systems, ethnicity, lifestyles, traffic patterns, etc.
I was intrigued to see how MapInfo provides this type of data, and more importantly, a set of graphical tools to analyze it. They showed me a map that a retailer could use in making new store expansion decisions. It showed hotspots indicating the best locations and estimates of annual revenue and profit for each new location. It provided several different scenario approaches and allowed you to play ‘what if’. The input data comes from the retailer’s own existing POS data as well as demographics from US census data, extrapolated for changes since the last census. MapInfo is religious about not commingling the POS data from different retailers, even anonymously. The census data does get stale. There are limits to the depth and accuracy of this and the POS data when drilling down into the “mind of the customer”. However, I think this approach really has legs.
Integrating Diverse Data
MapInfo links to other sources like LoopNet, the largest online real estate listing service, and partners with Accruent (Real Estate Lifecycle Management) to show current points of construction and lease obligations. These combinations of data sources can be used by a retailer to make smarter decisions, not just about where to locate a store, but how much space to allocate for each department and what mix of products are optimal for the surrounding and transient population.
MapInfo builds different analysis models to match the strategy of the retailer. For example, a “gravity model”, which focuses on population density and traffic patterns, supports a convenience-based strategy. I believe this is where the real value lies – taking a rich and diverse set of data, and building up over time a deep and rich set of knowledge, models, and analytic approaches that can be used to make sense of those data to answer the critical, difficult questions. I’ve updated the “Data Universe” diagram to reflect this.
Heard the Forecast?
Retailers always want to know, as far in advance as possible, what will my customer buy? A lot of it has to do with what is in the “mind of the customer”. But actual consumption, in virtually all retail categories, is also impacted significantly by the weather. The clothes we wear, the food we eat, the tools, gadgets and supplies we use, the entertainment we see (or don’t see)—they all change dramatically as the weather changes. Many retailers’ planning implicitly assumes that this year’s weather will be pretty much the same as last year’s. Living in New England, I can tell you it is virtually never the same as last year! Planalytics (another company I spoke with at NRF) provides long term, 12-18 month forecasts and tools to help retailers and other businesses analyze the impact for them.
They forecast the impact of weather on long range buying patterns, translating weather forecasts into advice about what it means for the retailer. Their computer models are updated every night against the reality on the ground, comparing what they predicted against what actually happened (I challenge the analyst community to be that brave—comparing every one of our forecasts against what actually happened!!). The weather forecasts and analytics are used to answer questions like how can I align promotions with the natural lift provided by the weather? How to more accurately forecast consumer demand, based on the weather, and make appropriate adjustments to allocation and replenishment plans? How to integrate promotional planning with merchandising activities ahead of the season? This is another example of taking a non-traditional data source and analytical approach to improving the quality of decisions.
The Human Element
Finally, I wanted to touch on two organizations I met at the NRF show that bring human beings into the equation, but in extremely different ways. The first is not a new concept – the Mystery Shopper. I spend a few elucidating moments with the Executive Director of the Mystery Shopping Provider Association. Done right, this can be used to really assess and improve the total shopping experience—everything from the attitude and knowledge of the associates, to the look and feel of the store, assortment and availability, convenience, etc. If the mystery shoppers are astute and articulate (and that is important), then you can get knowledge and insights that you could never obtain from a computer application or system.
I also spoke with the CEO of Applied Intelligence Systems (AIS). They realized that for certain jobs, such as assortment planning, there was a tremendous performance delta between the best vs. the average planner. Because people typically stayed in those jobs for a short time (e.g. 18 months), there is very limited opportunity for mentoring or passing on the skills. AIS created technology to analyze and capture the decision making process of the best performers and make it available to the masses. This is not another optimization tool. It works in conjunction with traditional optimization tools and gives suggestions, prioritized for likely success. I didn’t have time to validate how well this tool has worked in actual operation, but I found the concept fascinating: There are certain things that people do much better than computers (and vice versa). The idea is to capture the “way of thinking” and methods of the best performing people and make those available to the rest. The puzzle of integrating what humans do best with what computers do best, is a complex one. This may be one piece of the puzzle, and we’ll keep on eye on how it does.
Finding the Patterns
Our brains excel at recognizing patterns amidst a sea of sensory data. Those who didn’t recognize patterns fast enough got eaten by that saber tooth tiger! The examples here highlight ways that retailers can improve their “pattern recognition” capabilities – synthesizing data from very diverse sources like maps, weather, demographics, and the decision-making process of their best performers, to make much smarter decisions over time. Those companies that don’t get good at recognizing the patterns from this ocean of digital data are destined to become someone else’s lunch.
Click here to send questions or comments to the author.