Summarize the case as points And answer the two questions

Case

BIG DATA AND PREDICTIVE MARKETING

Big data refers to the deluge of digital data that is being produced by the billions of people using the Internet around the world, as well as an explosion of data from the Internet of Things. But big data is about more than volume; it is also about velocity (data comes in real-time torrents, loses value quickly, and requires rapid responses). variety (the data deluge contains both structured numeric data and unstructured data such as e-mail, video, and audio),. variability (the flow of data is event-driven and leads to peak loads, followed by relative calm), and complexity (the data comes from different sources and requires cleansing, matching, and reformatting in order to be useful).

Storing all this data requires new kinds of database technologies and analyzing it involves software called business analytics Big data can lead to better decisions and competitive advantages for firms that get it right and is influencing the design and marketing of retail products and in-store sales efforts. Big data and powerful analytics programs have given marketers the ability to send personalized messages to customers recommending products before they ask for them.

Predictive marketing is different from traditional in-person sales because it is based on the collection of data and the use of software to maximize the likelihood of a sale. Predictive marketing can scale to millions of customers and make decisions in milliseconds. Stitch Fix, an online clothing retailer using a monthly subscription revenue model, is one example of an online retailer using big data and predictive marketing. Stitch Fix blends expert styling advice, personalization software, and unique products to deliver an individualized shopping experience. New customers fill out an online Style Profile, which is then analyzed by the firms proprietary software to identify products that the customer is likely to purchase.

The company has thousands of personal stylists that interpret the output of the system and then handpick five clothing items and accessories each month that are unique to the customers taste, budget, and life style. The customer is not required to purchase the items until they have been received and accepted; the process to return the items is simple.

Over time, the software keeps track of what the customer purchased and learns to make better predictions based on what customers actually keep (as opposed to what they say they want, a key difference).

The more accurately Stitch Fix can predict what its customers will likely buy, the more sales it will generate. But using analytics to better understand its customers also allows Stitch Fix to reduce its inventory costs, to adjust production to better meet demand, to know its customers better than its competitors do, and even to fill and ship its orders in the most efficient way.

The data collected in the Style Profile includes basic demo-graphic information, plus a photo section that depicts seven different styles. Customers can respond to each style suggestion to further differentiate themselves in the Stitch Fix software, which runs on Amazon Simple Storage Service.

Customers can also share links to their Pinterest profiles to give Stitch Fix even more information. Based on the customers demographic information and selections of preferred styles, the software predicts which of several thousand products the customer would like.

Stitch Fix uses its customer data in a closed loop to make continual improvements to its come more accurate as they are exposed to more customer data. If tweaks to the algorithm are shown to be more predictive, they become permanent, whereas changes that fail to improve the algorithm are discarded.

The company has a team of 85 data scientists developing new tools and improving its core item selection algorithms. Eric Colson, the chief algorithms officer, joined the company from Netflix, whose recommendation engine is often cited as the gold standard in personalization. Stitch Fixs algorithms synthesize customer feedback, purchase and return decisions, and profile information to quickly generate possible recommendations.

Stitch Fix stylists use these results to make their next selections. For Stitch Fix, the blend of seven years of highly granular data, cutting-edge machine learning, and expert human input has been a winning combination. The company grew explosively throughout its first few years, reporting $977 million in revenues for the fiscal year ending in July 2018, up from just $73.2 million in 2014.

The company only raised $42 million in venture capital, a relatively small amount, and derives 100% of its revenues from subscriptions. Over three-quarters of its users describe them-selves as very satisfied or better.

In November 2017, Stitch Fix launched an initial public offering. Stitch Fixs profitability and solid growth were attractive to investors, and the stock has generally performed well. In the third quarter of 2018, Stitch Fix beat Wall Streets estimates for revenue and profit handily, sending the stock rising along with news of its launch of Stitch Fix Kids division.

However, in the fourth quarter, the stock missed its targets for user growth, and the stocks price fell again. For companies like Stitch Fix, investors want to see growth above all else, so the slowdown in user base raised concerns.

Amazon has also announced plans to launch a recommendation service that could eventually compete with Stitch Fix, further dampening enthusiasm for the stock. However, Stitch Fix has announced plans to expand into the UK to jumpstart its growth again, and as the company begins to spend more on marketing (it had relied almost exclusively on word-of-mouth from loyal users), it should once again see rapid growth in its user base, satisfying investors, even though the companys profitability may suffer slightly.

Traditional retailing companies are feeling the squeeze from both Amazon and Stitch Fix, and are trying to harness the power of machine learning and big data as quickly as they can. Macys and JCPenney have launched initiatives powered by machine learning such as in-store shopping assistants, and other brands have launched similar subscription services, such as Nordstroms Trunk Club. However, these companies lack Stitch Fixs expertise and experience in these areas and are likely to lag behind. Walmart has also thrown its considerable weight behind big data and predictive marketing to power the companys website and mobile app, and the aforementioned Amazon alsolooms as an omnipresent threat. Regardless of Stitch Fixs level of success going forward, the machine learning and predictive marketing algorithms that have powered its meteoric rise are likely to become the new normal in retail.

Case Questions

1. How does big data enable predictive marketing?

2. Are there any drawbacks to the increasing use of predictive marketing?