Xtreme Data Warehousing – Retailing Use Cases – 2

Ikea is a store that has done wonders in how they have you move through their stores, basically from one end to the other.  But a classic retailer doesn’t work that way.  Some stores have loyalty programs to try and do analysis on shopping patterns behaviours and some even do a good job of interacting with us in real-time (Travelocity is a great example to be discussed on another day).  So how can we leverage Xtreme DW to help us in this area

Total Shopping Experience Analysis
We are used to capturing the usual market basket whether we have identified the person or not, but this is only at the check out.  We miss the whole store experience up until that point.  So we are missing some interesting data that can be collected:

Market Basket – not just what is in the basket at checkout but

    • order and timing of placing items in and out of the basket
    • shopping route
    • video capture of the complete experience

So when we walk through an aisle and pick up the canned peas in one place, but get to fresh produce and put it down there, and then buy the fresh peas.  Or going from name brand to generic or vice versa.  Or going from smaller quantities to club packs.

Is it feasible (Technology yes and cost no respectively)

  • RFID on packaged goods, the technology is there, but currently too expensive
  • Digital Image Capture and Matching – yes it is available and being used at border control.  Most stores already have the cameras
  • Smart carts – go hand in hand with the RFID.  Yes the technology is there, but not at the right price point.

    What are the Issues

  • Big brother
  • Privacy

Let’s talk about another example in retail leveraging some of the above in a different manner which opens some moral questions.

Pricing and Promotion

  • competitive and markdown pricing
    • near-term transition to electronic shelf labels
    • pricing at the customer level through pricing and promotions)
  • Targeted pricing and promotions
    • based on the customer has identified to help you cross-sell

So as Dick and Jane walk through the store they are different
shoppers.  Jane is a regular and buys all her food except for fresh
produce.  Dick is a spuradic shopper.  They both walk up to the fresh
produce, Dick picks up a head of lettuce and it costs $3.00.  When Jane
picks up her head and checks the price it is $2.59.  Jane was
identified using our digital image capture and associated with her
history which showed that they were missing that part of the basket and
gave her a deal to try and incent her.  When Dick was captured, there
wasn’t any great history and he was given full price.  Another great
example is valentines day cards.  The time to start reducing the price
to eliminate the stock is gender specific.  Males buy there cards at
the last minute where women generally buy earlier.  Now do this in
real-time based on history and real live events.

Is it feasible – yes as before

  • RFID
  • Digital Image Capture
  • electronic shelf labels
  • electronic shopping cart with individual price checker

What are the Issues

  • Is it right that two shoppers at the same time get different
    prices.  What if we are neighbours and when talking about the great
    Turkey we had for Thanksgiving we realize that someone paid more than
    the other.
  • Is it any different than membership or loyalty rewards programs or couponing?
  • It works in corporate world where if I buy this from you and add
    this and now it is a significant purchase I expect a deal.  Or
    sometimes I reward for new customers and leave my existing customers to
    wallow – Cellular and High-speed internet are classic for this.

Leave a Reply

captcha *