We recently presented our work with Birchbox, the premier beauty subscription service, to a panel of judges for the INFORMS Wagner Prize for excellence in the practice of operations research and innovation in mathematics and advanced analytics. One judge asked how customer preferences and behavior are incorporated into the optimization model for product assortment.
We replied that Birchbox encourages the customer, after receiving a monthly box of products, to go to its website and report, “I liked this one” or “I didn't like that one.” This response helps Birchbox cluster similar subscribers.
The downstream app problem is important to the Birchbox operations executive. Subscribers are expecting a new box with new content every single month— as a discovery service, long-term subscribers tend to be ineligible for a lot of products because they have already received them. For a given month, let’s say you have 100 different boxes and configurations. You must buy and build 100 types of boxes. You might partner with a third-party logistics provider to assemble the boxes.
You will send 100,000 of Box 1 to a set of 100,000 subscribers, and 200,000 of Box 2 to another set of 200,000 subscribers. First, you must make sure every customer gets a box. Birchbox doesn’t stop there: the box must “excite” and engage the customer. Birchbox uses the word “excite” a lot. The goal is to let customers try and learn about new products through sampling: if they like the sample, they will buy a full-size product via Birchbox.
To measure this activity, Birchbox employs a proprietary methodology for measuring the excitement of their customers, as well as their likelihood to enjoy products and add them to their routine by purchasing the full size. Operations executives have reported an increase in excitement since they started with the new optimization model we developed for them. At Princeton Consultants, we have worked with countless KPIs—this marks the first time we have encountered an excitement score.
Client executives like to talk about KPIs, but they may not be able to crisply define them for the purpose of a mathematical model. A lot of business and stakeholder discussion may be necessary. Part of the art of optimization is defining a KPI, understanding how you will measure it, and collecting the required data at the right granularity. At Princeton Consultants we plan KPIs into the solution design and even start collecting them before the new system is fully developed.
Is “excitement” an appropriate KPI for your business? Consider this recent Glossy article, “Why Q4 is a Critical Time for Brands to Dive Into Subscriptions,” in which the authors note:
“The subscription business model delivers sustainable, long-term customer value. A study by Rakuten Intelligence highlights the dynamic: ‘Amongst the (mostly) pure subscription merchants (Birchbox, Dollar Shave Club and Ipsy), we see very high repeat customer rates, with 68, 82 and 81 percent, respectively, of customers’ purchasing more than three times.’ Since it costs more money to retain a customer than to acquire a new one, beauty brands can save money and reinvest those dollars by focusing on DTC [direct to customer] subscriptions. The subscription business model also empowers beauty brands to better understand their customers, uncovering valuable consumer behavior data, and giving shoppers access to essential items regularly, allowing them to save money through ‘subscribe and save’ deals.”
Understanding and engaging customers, driving activation and repeat sales—and measuring this process the Birchbox way. Now that’s exciting!
Contact us to discuss excitement and e-commerce optimization opportunities.