Is Gucci’s Fad Shopper Dilemma a Predictive Analytics Opportunity?
Gucci's runaway growth has largely stemmed from first-time, millennial customers – a notoriously fickle client demographic with little brand loyalty and tall customer experience demands. Will Gucci's hesitance to deploy machine learning be what proves it a fad, rather than an enduring fashion phenomenon?
Few industries have experienced a reckoning as brutal and relentless as that of high-end fashion. The accessibility of ecommerce and the affordability of fast fashion have rapidly democratized a business whose trend dictatorship once relied on elitism as much as it did quality design. Yet amidst this upheaval, one brand has leveraged sensational design to emerge an industry megastar: Gucci. Driving 9 quarters of double digit revenue growth has been a torrent of millennial shoppers – a notoriously fickle demographic with flimsy brand loyalty and tall customer experience demands.[1],[2] In this age of rapidly evolving consumer preferences, Gucci faces a daunting challenge: How to sustain momentum with a customer base loyal only to the next big thing?
Gucci’s millennial traction has primarily stemmed from product novelty – a dangerous strategy given fleeting trend cycles and fierce industry competition. Going forward, machine learning must become a critical business tool as the brand’s runaway new customer growth inevitably slows, and it increasingly relies on client retention to drive performance. Through amassing extensive customer databases, more and more savvy retailers have applied machine learning to predict future transactions and win customer loyalty. Such technology allows marketing professionals to forecast not only second purchase timing, but also products selected.[3] For example, customer data may reveal first-time shoe buyers typically purchase a handbag within 6 months of the first transaction. This can fuel highly effective client outreach campaigns, which improve marketing ROI and broaden a brand’s repeat customer base. The personal styling company Stitch Fix takes predictive analytics even one step further – to product design. By analyzing the purchase behavior of its customers, Stitch Fix identifies voids in its current merchandise selection and swiftly produces such designs themselves.[4] These efforts not only transition one-time shoppers to repeat clients, but also reduce the risk of fashion misses in a highly subjective industry.
While Gucci has taken preliminary steps to address data-driven opportunities, it has largely prioritized infrastructure above all else in the near-term. As a nod to its “digital competence”, L2 Research ranked Gucci the best performing digital fashion brand in 2016 and 2017 – across web, digital marketing, social media and mobile.[5] While L2’s scorecard primarily rates operational prowess, a meaningful digital presence still denotes extensive, omnichannel data collection – which powers comprehensive client insights and advanced customer analytics. Gucci anticipates additional IT and Customer Relationship Management (CRM) investments in the coming quarters.
In the medium-term, Gucci has also commenced a production cycle revamp, with the goal of accelerating customer demand reaction time. Based on its customer data findings, Gucci will likely look to manufacture popular styles and produce new designs more quickly.[6]
While these foundational investments are well-spent, Gucci must now rapidly move to address its analytics opportunity – and build a more predictable revenue stream grounded in known customer preferences rather than moonshot trend creation. As its first priority, Gucci should recognize those trends apparent in (what exists of) its repeat customer base. Predictive analytics can identify which products drove a second and third purchase – and Gucci can seek to replicate that behavior in first-time shoppers through personalized messaging and ad placement.
In the short- medium-term, Gucci should look to inform its design process with customer analytics. Historically, its brand leadership has balked at data-driven product design – fearing that data may poison the creative process. CEO Marco Bizzarri has said, “You need to start with creativity… it doesn’t start with numbers.”[7] Although the role of machines in trend creation is certainly contentious, the complete exclusion of analytics in design evolution ignores a vital customer development opportunity. To return to a previous example, if Gucci learns first-time shoe purchasers typically produce handbags as their second transaction, Gucci should invest in a broader accessory collection, across silhouettes and price points. This awareness of customer behavior allows the brand to not only predict purchase behavior down to the SKU-level, but to also ensure no product gaps exist in upcoming season collections.
If Gucci fails to retain its one-time, millennial customer base, it faces a material business risk. The value of analytics in customer targeting and merchandising strategy is tangible, and its utilization is ever more urgent in a business reliant on an unreliable clientele. With that said, however, the fashion industry’s view on machine learning and its role is largely split. On the one hand, we hear Marco Bizzarri asserting, “You need to start with creativity”[8]; and on the other hand, we see Gap transitioning to an entirely data-driven design model. Do sweeping customer retention strategies become moot if a company can repeatedly predict and command fashion trends? And at what point might machines cut out the creative process altogether?
(773 words)
[1] Kering Investor Presentations, October 2016 – October 2018, http://www.kering.com/en/finance/publications, accessed November 2018.
[2] Michael Osborne, “Brands Need To Step Up Their Game To Win Over Millennials,” CommunityVoice, Forbes.com, September 26, 2017, https://www.forbes.com/sites/forbesagencycouncil/2017/09/26/brands-need-to-step-up-their-game-to-win-over-millennials/#3a8f7a9f1b32, accessed November 2018.
[3] “Big Data Drives Luxury Brands Growth Beyond Digital,” February 12, 2018, Luxe Digital, https://luxe.digital/digital-luxury-reports/big-data-drives-luxury-brands-growth-beyond-digital/, accessed November 2018.
[4] Katrina Lake, “Stitch Fix’s CEO on selling personal style to the mass market,” Harvard Business Review Vol. 96, no. 3 (May/June 2018): 40.
[5] Alizah Farooqi, “Top 10 Fashion Brands in Digital,” Gartner L2, September 28, 2018, https://www.l2inc.com/daily-insights/top-10-fashion-brands-in-digital-4, accessed November 2018.
[6] Hilary Milnes, “To maintain growth momentum, Gucci is bringing production in-house,” Glossy, October 31, 2017, http://blog.else-corp.com/2017/10/to-maintain-growth-momentum-gucci-is-bringing-production-in-house-glossy/, accessed November 2018.
[7] Nikara Johns, “5 Surprising Things CEO Marco Bizzarri Had to Say About Gucci,” Footwear News, Yahoo, November 20, 2017, https://www.yahoo.com/lifestyle/5-surprising-things-ceo-marco-212842856.html, accessed November 2018.
[8] Ibid.
Great piece on a phenomenon highly relevant to our generation. While Gucci has definitely proven to be perhaps the most adept at finding success with our demographic, the question remains whether they can maintain their strong position even as the market passes the baton from millennials to Gen Z, a group that is already developing preferences that are different from the generation before it. Given that machine learning depends highly on the data it consumes, it is unclear whether the even Gucci’s analytics software will be able to capture shifts in trends if they occur too drastically. At the end of the day, I have to agree with Mr. Bizzarri that the creativity from human input is still needed, especially in the world of high fashion.
This article was an interesting take on the use of data in retail, since it is looking at the application of predictive technology in a section of the retail industry that is unpredictable. Unfortunately, I have to disagree that there is any possibility that technology will be able to predict trends and eliminate the creative process. The difference between Gucci and retailers who are currently using machine learning is that Gucci is setting trends, whereas other companies are learning to get as good at possible at knowing when customers will adopt trends. Amazon, Gap, and Stitch Fix’s success in the apparel industry is contingent on luxury brands like Gucci creating new trends that drive customers to buy more, even when their closets are full. To consider whether machine learning could replace the creative process in retail, I look other artistic industries (art, food, entertainment) and ask if could have their creative process replaced solely by data. Will artists, chefs, and movie producers become irrelevant in the future? I don’t think so.
I tend to agree with “Probably Alex” (LOL) in the above comment. I believe in the potential of machine learning and artificial intelligence but I also think that they have been overused as “buzzwords” in fashion and retail, just as Gucci can be a fad for those millennial shoppers. Data will be extremely influential in customer engagement (upselling new products, engaging churned customers etc), but data is also very backward looking. It helps to see what customers are ALREADY searching for, which means that those trends already exists. Customers can’t search for things that they don’t know that exists (sometimes they don’t even know what they want), so data may be not very useful to help guide trends. However, one caveat is that fashion doesn’t exist in its own isolation. It is impacted by pop culture and societal changes, usually with a lag. One example is the rise of streetwear in recent years, which can be explained by the popularity of hiphop that happened many years ago. Therefore, using machine learning and artificial intelligence, creative directors can detect and predict what pop culture trends are going to emerge and influence consumers. Those trends can help guide them with the actual creative process of fashion.
Great article! I feel that Gucci can leverage both creativity and data in their design process. I think a human driven approach makes most sense when trying to create an overall direction/perception of the brand, but analytics would be much stronger in deciding things like price points etc. so the company has to use both in order to succeed going forward