Box subscription companies are reinventing retail using personalization and artificial intelligence.

In recent years, amid a lackluster shopping environment, an unheralded retail phenomenon has taken off quite dramatically — with growth rates exceeding 1,000%. The startups representing this evolving segment are collectively grouped under the label of subscription e-commerce. With such evocative names as BarkBox, Birchbox, Blue Apron, Harry’s, OwlCrate, Trunk Club, and Winc, these online companies mail monthly boxes containing specially curated items in beauty, fashion, food, personal grooming, and pet products priced between $10 and $80 per box right to their subscribers’ doorsteps. Attention was brought to this category when consumer giant Unilever snapped up one of the best-known startups — Dollar Shave Club — in July 2016 for an eye-popping $1 billion — five times its annual revenue. That acquisition has drawn considerable interest to these companies, the reasons for their burgeoning popularity, and the tactics underlying their business practices.

A number of supply-side and demand-side reasons account for the success of box subscription companies. They have low startup costs, a direct-to-consumer framework, a lack of middlemen, a recurring revenue stream, and the potential to ramp up very quickly. For the time-pressed and budget-constrained consumer of today, these companies can provide tremendous convenience. Because the boxes are delivered directly to subscribers’ homes, they save the time and effort related to having to shop in stores. The added bonus is that subscribing to these boxes results in actual monetary savings for them: The monthly subscription fee is much less than what the items in each box are collectively worth. For the buyers there is an added element of mystery or delight in receiving boxes that are tailored to their individual tastes, and whose contents are not known beforehand — akin to a “Christmas every month” experience. For example, the most successful box subscription company, ipsy, founded by popular YouTube beauty guru Michelle Phan, sends its members personalized beauty boxes called Glam Bags for a monthly fee of only $10. These bags are very popular with the subscribers. Each one contains five deluxe makeup items from a wide range of brands, allowing them to try out different options every month at low risk and cost.

The role of data analytics is absolutely crucial to the success of the subscription e-commerce model. The biggest challenge that these companies face on a recurring basis is having to manage the rate of “churn,” or membership cancellations. If the box company can accurately predict what a customer will like — sometimes even better than what he or she is able to articulate — and then learn from his or her evolving tastes over time, then the probability of cancellation diminishes considerably. The AI-based recommender systems utilized by many of these companies allow the deployment of a preference elicitation framework, initially done by customer survey. The decision software then mines the purchase history to make future recommendations. Most companies require their customers to fill out a data profile stating their personal preferences. Ipsy gives its first-time members a “beauty quiz” asking about such things as skin tone, eye and hair color, and favorite brands, so that each Glam Bag may be useful — there are more than 200 possible configurations of Glam Bags monthly.

Perhaps the best exemplar of a box subscription company making full use of AI-based data analytics is the “personal styling service” Stitch Fix. This San Francisco-based company, founded in 2011, has thousands of trained personal stylists who recommend clothing choices for women (and now men) who are either too busy to shop or who are not too familiar with fashion. A box containing five curated items of apparel and accessories, termed a “Fix,” is shipped monthly to customers. They keep the items they like and return the ones they don’t. What differentiates this company from its competitors is that it employs a large team of some 60 data scientists utilizing AI-based machine-learning algorithms to assist the stylists in selecting only those things that clients are likely to keep. When prospective customers sign up for the service, they are asked to fill out a fairly detailed “style profile” that asks about their personality, body type, lifestyle, budget, favorite styles, colors, and even what bodily features they would like to highlight or downplay. Then the style forecasting algorithms go to work, churning out the best matches for clients. Other AI algorithms keep track of a customer’s satisfaction with the service over time and how likely they are to renew or drop out. The most interesting aspect about these algorithms is that they learn from experience, get smarter after every purchase or non-purchase, and make even better predictions with time. As a result, clients find the service has become their indispensable “fashion BFF” that seems to understand their needs better than they do themselves.


The final piece of the puzzle underlying the success of these companies is their effective use of social media and influencer marketing. Many box subscription companies have organized their customers into online communities where they can offer useful feedback, share product suggestions, assist new users, and importantly, help recruit new members. NatureBox, an online subscription startup delivering monthly boxes of healthy snacks, offers a good case in point. It sells snacks from over 100 possible choices including vegan, dairy-free, nut-free, non-GMO, and soy-free options. Subscribers get five snacks in each box that have been approved by a nutritionist. Social media marketing is very important to NatureBox in acquiring new customers and retaining existing ones. With over 1 million followers on various sites like Twitter, Facebook, Instagram, and Pinterest, the company engages customers in contests, recipes, and conversations with one another. It was reported by USA Today that some 50% of NatureBox members share their experiences on Facebook.

When studied closely, the tactics implemented by box subscription services are reminiscent of the wildly successful Spanish fast-fashion chain Zara. Its business model is based on total vertical integration from manufacturing to retailing. This allows Zara to maximize the items sold at regular prices within a time window by bringing fashion from concept to store much quicker than its competitors. By extension, the successful box startups like ipsy, graze, and Stitch Fix that are implementing the “better, faster, cheaper” model can be said to be pushing the “Zara in a box” idea. Not surprisingly, many young entrepreneurs and savvy venture capitalists have been attracted to this evolving market in the United States and elsewhere. Lately, these services have also penetrated the European market, with companies such as Fairybox, GLOSSYBOX, and Pink Box. Even industry giants such as Procter & Gamble, Walmart, Starbucks, Adidas, Macy’s, and Nordstrom have joined in, either via acquisition or by starting their own online subscription outlets.

For businesses wanting to enter this space, four guidelines may be given here.

First, choose the right product category. Box subscriptions are broadly well-suited for two category types: those that need to be replenished on a regular basis such as food, pet care products, health supplements, or personal grooming products, and those in the “self-gifting” category like books, jewelry, cigars, wine, and so on. Second, focus on personalization by implementing the latest supply-chain technology and data analytics. Third, provide a high level of service at the individual client level. The tactics adopted by the successful box services hark back to a more genteel, client-friendly era. And finally, foster a close sense of community among the members using social media.

1 Comment On: How Analytics and AI Are Driving the Subscription E-Commerce Phenomenon

  • richard espinosa | December 22, 2016

    Nice article I must say. Even my friends enjoyed reading as we’re researching for assignment presentation. Glad this helped us so thankyou.

Add a comment