Predictive Analytics Times
Predictive Analytics Times
EXCLUSIVE HIGHLIGHTS
SHARE THIS:

1 year ago
The Evolving State of Retail Analytics in CRM

 

The Traditional State

The world of retail has undergone a revolution on the use of data for CRM analytics. Historically, data was simply unavailable for CRM analysis unless the retailer had a loyalty card program which essentially was the data capture mechanism. Certainly, much has been written about the tremendous value of loyalty programs and the opportunity in leveraging its rich information for more profitable decisions. But the vast majority of retailers operate their business without any loyalty program. The only data capture mechanism was the point of sale terminal which was very limited due to its inability in capturing information which can be attributed back to the individual customer. For these retailers, the level of any detailed analytics was confined to the store level. Retailers can certainly identify their best stores and worst stores from this information. At the same time, profiles could be developed around these type of stores. For example, what do my best stores look like? As a store owner, I might want to know the following:

  1. What is the product mix of my inventory?
  2. What are the payment vehicles that are used by my customers?
  3. What do the prospects look like that live close to these stores?

The first two bullet points can be collected directly from store information while the third bullet point would typically be gathered by a third party provider. Because the critical information concerning the prospect pertains to where he or she lives relative to the store, the immediate task is to define prospects that live within a certain trade area of that store. The trade area is not always a precise definition. In some cases, this trade area could be 2 miles, five miles, or 10 miles around that store and of course will depend on the nature of the business. With the prospects now having been defined, our next step is to capture information based on the geography of where they live such as:

  • Income
  • Education
  • Wealth/investments
  • Immigration
  • Purchase habits
  • Social media/digital habits
  • Media consumption

The richness of this above information can be used to develop more detailed type profiles on the makeup of prospects who live within the trading area of these stores. A number of different store strategies can then be adopted based on these profiles:

  • Poorer performing stores need to adopt marketing strategies that will attract customers which fit a given store prospect profile.
  • Best performing stores need to continuously align their product inventory strategies which is consistent based on the prospect profile of those stores.
  • Certain stores will be closed since there is a limited opportunity of attracting customers as the store profile of these prospects is too dissimilar from the best stores’ prospect profile.
  • New stores and their locations will be determined based on the fact that they most resemble the prospect profiles of the best stores.

Using Mobile Technology

However, the analytics is still restrictive in the sense that the retailer cannot execute targeted market programs directly to that consumer unless as stated above there is some type of loyalty program. But this has all changed now with mobile technology. With mobile technology, the phone itself becomes the unique identifier for that individual. As long as there is some sort of router within the retail establishment, the technology can now capture the data that is being transmitted between the phone and the router presuming the individual’s cell phone is turned on. This sensor data can then be converted to the following:

  • Time of entry into establishment
  • Time of exit from establishment
  • Distance from router
  • Store ID

Think about these four pieces of information as we can now derive rich variables pertaining to duration, time of day, time of week, as well as information related to recency and frequency of store visits. All kinds of change variables can be created as we now have information related to time. Since store information is available, we can create customer variables related to store usage. The variable derivation possibilities are limitless but an upfront data audit would again be a key tool in determining what possibilities make the most sense. Predictive models can be built based on the likelihood of that customer returning to the store within a defined period of time. Marketers can then leverage this information to market directly to this individual on his phone either through an in-personalized SMS text message or more personal text message presuming the customer has opted in through some login type platform.

The use of beacon type technology provides an accelerated version in capturing information about individuals. Using the individual’s location in the retail establishment, triangulation of the sensor data between the cell phone and the router can now indicate distance between the individual and particular items in the store. This information can then be used to communicate certain messages to the consumer based on his or her location within the store. Yet, the real home-run is to create predictive analytics solutions based on the individual’s navigation history in the store. The creation of variables related to the historical navigation patterns through the store can be used to develop insights about that individual’s potential shopping behavior. Once again, mobile marketing techniques can leverage this information and communicate directly to the customer as he or she navigates the store. But the difference here in using predictive analytics solutions is that marketing is basing its communication decisions on what the consumer is expected to do rather than on simply where the customer is within the store.

It is still early days and certainly retailers can communicate to you based on your current location. But the ability to use all the history of the individual’s navigation patterns and to incorporate this information alongside the customer’s current location is not a common practice among st retailers. Yet, this is the real advantage for retailers in building predictive analytics solutions. The ability to predict behavior based on all the navigation location history rather than just the current location history represents the next challenge for retailers but certainly a worthwhile one as businesses all understand the tremendous value of predictive analytics solutions.

Author Bio:

Richard Boire, B.Sc. (McGill), MBA (Concordia), is the founding partner at the Boire Filler Group, a nationally recognized expert in the database and data analytical industry and is among the top experts in this field in Canada, with unique expertise and background experience. Boire Filler Group  was recently acquired by Environics Analytics where I am currently senior vice-president.

Mr. Boire’s mathematical and technical expertise is complimented by experience working at and with clients who work in the B2C and B2B environments. He previously worked at and with Clients such as: Reader’s Digest, American Express, Loyalty Group, and Petro-Canada among many to establish his top notch credentials.

After 12 years of progressive data mining and analytical experience, Mr. Boire established his own consulting company – Boire Direct Marketing in 1994. He writes numerous articles for industry publications, is a well-sought after speaker on data mining, and works closely with the Canadian Marketing Association on a number of areas including Education and the Database and Technology councils. He is currently the Chair of Predictive Analytics World Toronto.

Leave a Reply