Fundamental Models in Data Science

1. Classification (probability estimation or scoring): binary or categorical.Attempt to predict, for each individual in the population, which of a (small) set of classes this individual belongs to. Classification will bucket individuals, and scoring will provide quantification of likelihood of being in a particular bucket.

2. Regression (“value estimation”): numeric. attempts to estimate or predict, for each individual, the numerical value of some variable for that individual.

3. Similarity Matching: attempts to identify similar individuals based on data known about them.

4. Clustering: attempts to group individuals in a population together by their similarity, but not driven by any specific purpose.

5. Co-occurrence (also known as – frequent item mining, association rule discovery, and market-basket analysis): attempts to find associations between entities based on transactions involving them.

6. Profiling (also known as behavior description): attempts to characterize the typical behavior of an individual, group, or population.

7. Link prediction: attempts to predict connections between data items, usually suggesting that a link should exist, and possibly also estimating the strength of the link.

8. Data reduction: attempts to take a large set of data and replace it with a smaller set of data that contains much of the important information in the larger set.

9. Casual modeling: attempts to help us understand what events or actions actually influence others.

Should Zappos take steps into the hospitality world?

“When people talk about Zappos, it’s not just about the great pair of shoes they scored, but the awesome customer service they received. It truly has become a tangible asset that is synonymous with the Zappos brand. In extending the brand to hotels and beyond, the customer satisfaction bar will be set high. Execution will be key to protecting the brand’s equity. Also I could see some Amazon (the parent company) products and services being a part of this.” ~Shawn Harris

Read full article:


Saks, Ralph Lauren lure customers with upscale services.

“As discrete service offerings, both Ralph Lauren’s “taxi” and Saks’ merchandise delivery service would be of value to shoppers, as they should help to save a shopper’s time or otherwise bail them out in a time of need. However, they must be a part of a coherent strategy of driving footfalls that includes digital engagement. These services should feel like a natural fit to the overall experience, as opposed to feeling like a disjointed one-off.” ~ Shawn Harris

Read full article:

8 Experts Predict the 2016 Holiday Shopping Season #Retail #Holiday2016

Lean thinking

Shawn Harris, N.A. Retail & Hospitality Industry Lead, Zebra Technologies: Retailers should look at last year as a turning point, where more shoppers chose online vs. offline on Black Friday 2015. This means they should go leaner on in-store inventory (don’t worry, the majority of shoppers will start with your website anyway), ensure their websites can truly dynamically scale to meet load, use broadcast media to create theater and excitement in combination with personalized targeted ads on social platforms, and deliver what they say they will deliver.”

Full article:

Bed Bath & Beyond’s Membership Model #retail #marketing #membership

Is this Everyday Low Price (EDLP) spun as a membership program?  Will the elimination 0f the 20% Coupon’s “scarcity,” remove BBBY’s greatest customer activation trigger, or will the $29 annual fee create the “shop here first” behavior that Amazon enjoys with its Prime Members? The Amazon Prime comparisons always gets me as none of these membership programs come with all of the other value-adds that Prime does, ( streaming, cheap unlimited music). To truly compete, I think more complementary partnerships are required to enhance the value of these initiatives (e.g. discounted Uber/Lyft rides, services, etc [requested through the BBBY app] ), plus the extension of this membership’s benefits to the other BBBY subsidiaries. All in all, I do like BBBY’s willingness to test bold initiatives like this.

Reference Articles:

AmazonFresh lowers annual subscription via a $15 monthly rate. #Retail #Grocery

“I recall one of my first lessons in grad school, “cash flow is king.” I think that a lot of people feel the same about how they manage their home finances. Though there is only a $20 annual net savings, I think that the $14.99 per month fee will significantly lower the barrier to adoption for many Prime members. I think that this pricing move, and the recently announced push for both AmazonFresh pickup locations and perishables-focused convenience stores will position them to continue to grow their online grocery business and grow share.

Though the online grocery industry is relatively small at $33 billion as compared to the grocery industry as a whole at $795 billion, Amazon currently represents 26 percent of the online share with $8.7 billion in revenue. Convenience and competitive pricing will continue to reign … save me time, save me money.”  ~ Shawn Harris

Read full article:

Will customers give Amazon the keys to their smart homes?

“I completely believe that this is a concept that could see wide adoption. Airbnb has helped in resetting the idea of what personal space means and blockchain technology will allow for secure, immutable, one-time access to home IoT locks. Delivery person tracking and home tracking (cameras, mobile device and presence sensors) will play an over-the-top role for auditing behavior. Insurance will cover the rest.”  ~ Shawn Harris

Read full article:

Consortium For Operational Excellence In Retailing (COER) @Wharton – Day 2 Quick Recap

About Consortium For Operational Excellence In Retailing (COER)

Consortium for Operational Excellence in Retailing (COER) is focused on advancing retail operations from a combined academic and business perspective. We hold an annual conference in May, alternating between Harvard Business School and The Wharton School, where we present cutting edge academic research for participants to exchange ideas, thoughts, and challenges. COER attracts companies and academics from various parts of the world.

COER began as the Harvard/Wharton Merchandising Effectiveness Project in 1996, started by Marshall Fisher of The Wharton School and Ananth Raman of Harvard Business School. The academics in COER have published dozens of papers in leading journals and many case studies that are taught at top business school. The work produced by COER was summarized recently by Fisher and Raman in the book “The New Science of Retailing,” Harvard Business School Press. COER has facilitated the work of numerous doctoral students, many of whom currently are on the faculties of leading business schools.

COER grew out of the understanding that while the retail industry now has the analytical tools to make merchandising more effective, there are still many areas where academia can help to push the retail industry forward from an operational perspective.

Consortium For Operational Excellence In Retailing (COER) @Wharton – Day 2 Quick Recap

Session Eleven: Impact of Stockouts
Presentation by Ananth Raman
Key takeaways:
* In manufacturing, a 1% increase in historical in-stock is associated with a 12% increase in demand.w
* If you have a supplier that delivers and one that doesn’t, you’ll order more from the one that does, even if at a slightly higher cost. It’s like buying insurance.

Session Twelve: Consequences of Centralizing Hiring at a Retail Chain
Presentation by Tatiana Sandino of Harvard Business School
Key takeaways:
* Employee Departures:
** Centralized hiring results in a lower rate of employee departure in more busy stores.
** Centralized hiring results in a higher rate of employee departures when the store serves service‐sensitive customers.

* Store Performance:
**Centralized hiring is associated with greater sales in distant stores: 1% increase in sales/additional 10 miles away from HQ
** Centralized hiring is associated with lower performance where customer relations may be important:
*** 7.3% decrease in sales if store serves service‐sensitive customers.
*** 0.04 point decrease in (0‐1) mystery shopper score scale when store serves repeat customers.

Session Thirteen: Demand During Store Liquidation: Local Economic Factors
Presentation by Nathan Craig of Ohio State University
Key takeaways:
* Across retail segments, revenue and asset recovery rates during store liquidation are positively associated with local median household income
* Revenue and asset recovery rates are negatively associated with local consumer sentiment
* Initial inventory, last-year revenue, local median household income, local consumer sentiment, and chain effects explain much of the variance in liquidation revenue (R2 = 0.96)

Session Fourteen: How Retailers Respond to Demand Shocks
Presentation by Vishal Gaur
Key takeaways:
* High inventory turn (HIT) retailers are able to react much more quickly than low inventory turn (LIT) retailers. Margin impact is greater for LIT, than HIT.
* HIT retailers lever is quantity changes, LIT is price changes. HIT retailer have better sustained return on assets in shocks than LIT.
* …so tune inventory for turn, vs availability.

Session Fifteen: Consortium on Patient Experience (COPE)
Presentation by Ananth Raman
Key takeaways:
* Replicate COER for patient experience. More to come…

Session Sixteen: Spatial Competition and Preemptive Entry
Presentation by Fanyin Zheng of Columbia Business School
Key takeaways:
* Deciding on store location, based on the future entry of competition.

Session Sixteen: Using Peer Feedback in Performance Evaluation
Presentation by Serena Loftus of Tulane University
Key takeaways:
* Multi-source feedback has advantages over single source
** Manager note always working w/ subs
** Manager bad evaluation
** Bias of manager.

Session Seventeen: Mobile Technology location-based marketing in Retail
Presentation by Jose Guajardo of Haas School of Business, UC Berkeley
Key takeaways:
* Effectiveness: Non-GEO < GEO < Facebook * Distribution: Non-GEO > GEO > Facebook