02 Dec 2008

…brick-and-mortar book stores …: Part 3 — brick&mortar advantage

Raj Bhatt

As I had discussed in Part 1, Barnes & Noble and Borders need to consider three initiatives to compete against Amazon (in addition to shutting down unprofitable retail stores and reducing employee count):

  • Create a more credible online option to compete against Amazon
  • Use brick-and-mortar as a competitive advantage
  • Align inventory to demand

In this post, I shall focus on how B&N and Borders can use brick-and-mortar as a competitive advantage and better align inventory to demand:

  • Offer the option to buy online and pickup/deliver from a store : This is a feature that BN and Borders can uniquely offer, which Amazon cannot replicate. BN and Borders currently offer free pickup at stores. This needs to be promoted much more (while they still have many stores within driving distance of most of the US population). BN and Borders should also consider using their book stores as local delivery hubs to dramatically reduce their courier costs.
  • Better organize in-store book discussions/signings/ reading sessions: While both BN and Borders organize in-store events, i believe they do not use their facilities to the maximum. Borders does not promote in-store events on their website. Although BN does promote events on their website, they do not have a mailing list for events. Store associates should be encouraged to promote events and signing up to the event mailing list. BN and Borders could leverage their loyal customer base by allowing customer groups to ‘self-organize’ reading sessions at the stores, especially for children. Create a whole new group of ‘book moms’ among the loyal customers. Amazon admits it cannot compete with BN and Borders on in-store events !
  • Offer bundled offers in the store: Amazon scores over BN and Borders by offering deals on bundles of products based on prior transaction history. This is done using an analytical technique called Market basket analysis. There is no reason why BN and Borders cannot offer bundled deals in the store and online.
  • Install digital displays in the store showing reader reviews and suggesting complementary reading recommendations: Take collaborative filtering to the brick-and-mortar world by installing digital displays or dumb interactive terminals showing reviews for books in an aisle and recommending complementary books based on collaborative filtering. The CF algorithm should not be limited to the online sales data of BN/ Borders but should include offline sales also.
  • Here’s an outrageous marketing suggestion: pilot a reading room service. For too long many visitors to book stores have been using them as free reading rooms. They would not mind paying a few dollars an hour for a comfortable reading room.
  • Better align inventory to demand
  1. Use statistical models based on demographic profiles of store localities and historical ‘comparable’ title sales to better allocate new title inventory to stores. I bet that the sales of books related to hunting and the outdoors sell much more in the countryside and that sales of financial and management books are higher in urban areas.
  2. Dynamically trans-ship best-seller inventory between stores based on matching with demand patterns. This is done routinely for new DVD sales on a daily basis by Warner and Universal. It is noteworthy that most demand for new best-sellers occurs in the first week after release!
  3. Since BN carries more than 4 months of inventory and Borders carries more than 5 months of inventory, I can’t help thinking that they need to do better at demand forecasting even for mature titles. They should consider panel regression techniques which consider store characteristics, locality demographics and product characteristics (genre, author, rating, time since release, price, etc).

As a fan of Barnes & Noble and Borders book stores, I hope that these measures help resurrect these companies. I would love to hear about your thoughts.

brick-and-mortar, demand, forecasting, inventory, market basket analysis, panel regression, statistical models

Add

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>