SEPHORA: Developments in Machine Learning and Augmented Reality towards customer personalization.

Sephora as the leader in technological innovation within the retail industry has made significant advances in using data science to advance product personalization and customization. This essay surveys recent developments and poses questions to the challenges in inventory management and production technology the personalization trend entails.

With the advent of e-commerce many predicted an end to traditional Brick & Morter stores[1]. As the CEO of Stichfix, Katrina Lake, indicates, the disappearance of Blockbuster in favor of digital rival Netflix made many question their sales strategy[2]. However, recent data seems to suggest Brick & Mortar is an essential part of the consumer experience, and that the offline and online must coalesce to provide the personalized service customers demand[3][4][5].

Studies suggest machine learning can help unify these online and offline experiences through the analysis of  individual preferences and characteristics[6]. Sephora is the leader in retail focused on this trend as we see in their recent introduction of the ‘virtual artist’.

THE VIRTUAL ARTIST

Their virtual artist employs machine learning techniques to develop unique recommendations for users. Customers upload a base picture of themselves to ‘try products on’. Using visual data-points and Sephora’s costumer data, the Virtual artist recreates the said photograph with modifications showing the effect of the selected (or recommended) make-up.  Customers may try different products until they find their preferred choice. This improves and simplifies the purchase experience (see Figure 1), as we see reflected in the recent sales increase (+13% for the Selective Retail area of Sephora’s mother company LVMH[7])  and repeat purchases.

The Virtual Artist not only works towards increasing sales, it aims to increase engagement with consumers. Millennials and Generation Z consumers show a higher appraisal of engagement with companies through the customer experience[8]. To improve customer engagement the virtual artist is integrated with Sephora’s mobile app and loyalty program. As data is analyzed by machine learning techniques product recommendations improve and loyalty is seen to increase[9].

Thus, through personalization Sephora has solved the two key issues that impact the retail business today:

  • Bridging the offline and online experiences to grow customer loyalty
  • Driving sales in a highly competitive market where brand, product and channel options have dramatically increased[10].

SEPHORA: THE ROAD AHEAD

The virtual artist is one element within Sephora’s strategy of providing customers with a number of tech applications that enables them to personalize their shopping experience. Long term they aim to eliminate what Mary Beth Laughton (Sephora’s vice president for omni retail) defines as pain points[11]: the parts in the customer journey that end the customer experience before purchase.

Sephora innovates via their innovation lab. Here they experiment with new technological developments such as the color shade-matching technology called ColorIQ.  It is the hope that through these initiatives “If you walk into a Sephora five or ten years from now, every aspect of your experience, from what you see to the products you are recommended will be customized based on your face shape, your sales history, and your preferences”[12]

IMPROVEMENTS TO STRATEGY

It is the opinion of the writer that to achieve this Sephora’s efforts on customer journey and experience personalization need to be complimented by 1) improved inventory management processes and 2) flexible product development lines.

Today inventory management poses a problem for most retailers. Predicting sales on a large number of SKU’s in the context of rapidly changing customer behavior is a challenge that most retailers are working hard at solving. Sephora’s efforts in this field are remarkable, but, will have to improve if they are to hit their complete personalization target. With an increase in personalization tactics and offerings comes an increase in product customization, and consequently in number of SKUs. Because machine learning can aid in the process of predicting inventory usage I would encourage Sephora to make this another pillar on their personalization strategy.

Another element that Sephora will need to focus on to deliver this complete personalization is operations technology. To achieve the high degree of customization that ‘a product per customer’ implies, Sephora will require technology that can produce a large assortment of products in small lots at the current price point. Since reduction in product lots can drive costs up, new technological developments to keep costs down are essential.

In brief, Sephora is the leader in personalization innovation and has demonstrated considerable prowess to deliver this. Their focus on personalization, if effective, will lead to an increase in sales and customer loyalty. To help their efforts Sephora will need back-end solutions that complement the front-end innovations represented by the new virtual artist.

UNADDRESSED ISSUES

Beyond the above defined issues, there are some questions in the quest for customization and personalization of consumer goods that are yet unanswered:

  • Predictive analytics naturally leads to the standardization of products. Marketers and business leaders need to think of how to disrupt the market while relying on twenty-first century data analytics.
  • As people demand faster and more innovative delivery, cost control and scale within customization are key concerns.

Word count: 788

Bibliography:

  1. Adtaxi “2018 Online Shopping and Technology Trends Survey” [https://www.adtaxi.com/blog-roll/2018/6/12/2018-online-shopping-and-technology-trends-survey] Accessed November 2018
  2. Brynjolfsson and A. McAfee. ”What’s driving the machine learning explosion?” Harvard Business Review Digital Articles (July 18, 2017).
  3. Alison Delisco Rayome, “How Sephora is leveraging AR and AI to transform retail and help customers buy cosmetics”, Techrepublic, [https://www.techrepublic.com/article/how-sephora-is-leveraging-ar-and-ai-to-transform-retail-and-help-customers-buy-cosmetics/] Accessed November 2018
  4. Fedyk, “How to tell if machine learning can solve your business problem” Harvard Business Review Digital Articles (November 15, 2016).
  5. Krista Garcia, “Offline Retail Is Rebalancing”, emarketer pro, September 5 2018 [https://content-na1.emarketer.com/offline-retail-is-rebalancing] Accessed November 18
  6. Daniel Keyes, “Retailers are still lacking on the personalization front”, Business Insider Intelligence, August 10 2017 [https://intelligence.businessinsider.com/post/retailers-are-still-lacking-on-the-personalization-front-2017-8] Accessed November 2018
  7. Daniel Keyes, “DIGITAL COMMERCE AND GEN Z: How retailers and brands can reach and appeal to the next generation of consumers”, Business insider Intelligence, December 21 2017 [https://www.businessinsider.com/the-digital-commerce-and-gen-z-report-2017-12] Access November 2018
  8. Lake, “Stitch Fix’s CEO on selling personal style to the mass market”, Harvard Business Review 96, no. 3 (May/June 2018): 35-40.
  9. LVMH, “2017 Record Results”, January 25 2018
  10. Mintel, “Executive Summary Beauty Retailing, US September 2018” September 2018.
  11. Felix Richter, “Who’s Surviving the “Retail Apocalypse”?”, Statista, April 17 2018 [https://www.statista.com/chart/13550/change-in-retail-store-count-by-category/] Accessed November 2018
  12. J. Wilson, S. Sachdev, and A. Alter, “How companies are using machine learning to get faster and more efficient” Harvard Business Review Digital Articles (May 3, 2016).
  13. J. Wilson, A. Alter, and S. Sachdev. “Business processes are learning to hack themselves” Harvard Business Review Digital Articles (June 27, 2016).
  14. Yeomans. “What every manager should know about machine learning” Harvard Business Review Digital Articles (July 7, 2015)

Footnotes: 

[1] Felix Richter, “Who’s Surviving the “Retail Apocalypse”?”, Statista, April 17 2018 [https://www.statista.com/chart/13550/change-in-retail-store-count-by-category/] Accessed November 2018

[2] Lake. “Stitch Fix’s CEO on selling personal style to the mass market”, Harvard Business Review 96, no. 3 (May/June 2018): 35-40

[3] Daniel Keyes, “DIGITAL COMMERCE AND GEN Z: How retailers and brands can reach and appeal to the next generation of consumers”, Business insider Intelligence, December 21 2017 [https://www.businessinsider.com/the-digital-commerce-and-gen-z-report-2017-12] Access November 2018

[4] Adtaxi “2018 Online Shopping and Technology Trends Survey” [https://www.adtaxi.com/blog-roll/2018/6/12/2018-online-shopping-and-technology-trends-survey] Accessed November 2018

[5] Krista Garcia, “Offline Retail Is Rebalancing”, emarketer pro, September 5 2018 [https://content-na1.emarketer.com/offline-retail-is-rebalancing] Accessed November 18

[6]Daniel Keyes, “Retailers are still lacking on the personalization front”, Business Insider Intelligence, August 10 2017 [https://intelligence.businessinsider.com/post/retailers-are-still-lacking-on-the-personalization-front-2017-8] Accessed November 2018

[7]LVMH, “2017 Record Results”, January 25 2018

[8] Daniel Keyes, “DIGITAL COMMERCE AND GEN Z: How retailers and brands can reach and appeal to the next generation of consumers”, Business insider Intelligence, December 21 2017 [https://www.businessinsider.com/the-digital-commerce-and-gen-z-report-2017-12] Access November 2018

[9] A measure of customer loyalty improvement is the increase in repeat purchases previously described.

[10]Mintel, “Executive Summary Beauty Retailing, US September 2018” September 2018.

[11] Alison Delisco Rayome, “How Sephora is leveraging AR and AI to transform retail and help customers buy cosmetics”, Techrepublic, [https://www.techrepublic.com/article/how-sephora-is-leveraging-ar-and-ai-to-transform-retail-and-help-customers-buy-cosmetics/] Accessed November 2018

[12] Alison Delisco Rayome, “How Sephora is leveraging AR and AI to transform retail and help customers buy cosmetics”, Techrepublic, [https://www.techrepublic.com/article/how-sephora-is-leveraging-ar-and-ai-to-transform-retail-and-help-customers-buy-cosmetics/] Accessed November 2018

Previous:

The NFL and Machine Learning: A Touchdown for Technology?

Next:

The Benefits of Additive Manufacturing at BIA

Student comments on SEPHORA: Developments in Machine Learning and Augmented Reality towards customer personalization.

  1. This is very interesting, especially given the role of Sephora in the value chain. As a retailer and not a manufacturer of majority the make up products that are inputs into its algorithms, I wonder whether there are further concerns that needs to be highlighted. How does the frequency of listing (launching) a new product and un-listing a product affect the learning capabilities of a machine. Also, given that the definition of beauty has evolved over time and seems to be evolving what is the merit of leveraging machine learning that uses largely historical data and how can a machine learn beauty trends that seem to be socio-cultural and dynamic.

  2. This piece raises an interesting question about brand management and new product launches. If choices are driven by recommendation engines, how are products prioritized in that funnel? Given Sephora’s power in the channel, will their recommendation engines be biased towards products / companies that offer Sephora better terms?

    I think there is huge value in the creation of this data asset as well — not only can Sephora use this data in order to better develop their own product lines, but they could sell this data to makeup brands so that they may adjust their product development.

Leave a comment