How much is customer insight worth?
How about more than $1 million in sales for a single product? That's how much L.L. Bean Inc. earned last year by selling an Oxford shirt that the apparel and outdoor equipment company had pulled from its catalogue-until a surge of customer emails and reviews requested that the item be made available again. Yet, Bean Senior Vice President and Chief Marketing Officer Steve Fuller points out that customer insight is not worth much on its own.
"Customer conversations are truly a means to an ends," Fuller asserts. "The power of the customer conversation is the actions you take in response to it."
More companies are tapping this power source by gathering customer insight from a growing number of channels, including social media platforms, live-chat tools, Web forms, customer ratings and reviews, website analytics, sales conversations, call center discussions, and much more. L.L. Bean recently placed a computer screen next to a hiking shoe display in its Freeport retail store, for example. When customers pick up a boot and walk by the screen, a radio frequency identification (RFID) tag embedded in the display shoes' soles triggers product information, as well as customer ratings and reviews, about the specific shoe to appear on the screen.
"We can see which shoes customers are picking up," Fuller notes, "and how long they're holding each shoe." That will soon translate to even more customer insight-if Fuller's team takes the right actions. This is the big "if" companies contend with as they try to gather and act on more customer insight. Identifying the most profitable actions to take requires a lot of testing, data-crunching, and time, as well as clearly defined processes and a human touch.
Two essentials: time and data
The value of the insight depends on several factors, including historical trends, consumer engagement levels, conversion rates, and website analytics. "[Our] most valuable customer insight probably comes from extrapolating data from our website analytics software and aggregating information from data captured through our Web forms and live chat tools," explains Tej Shah, director of e-commerce and marketing for Blue Soda Promo.
Shah describes multivariate A/B testing as a "crucial" step in distinguishing between signals and noise. When Shah and his team's testing and analysis produce the former, they present these insights to their colleagues in sales, marketing, operations, or fulfillment who are best positioned to take actions that lead to improvements in sales and service. For example, in recent months Shah and his team conducted several tests on how to gain the greatest improvements in lead capture and sales. They deployed Web forms and Bold Software's live chat tool on Blue Soda Promo's e-commerce site to do so.
By making live chat available throughout the website and increasing the number of "contact us" forms available on the site, Shah reports that Blue Soda promo increased its sales conversion rate by 35 percent. Shah discovered that timing matters; potential customers who received marketing emails when they were further along in their decision-making process boasted 20 percent higher conversion rates than potential customers who received similar marketing emails earlier in their decision-making process. Similarly, by inserting a contact-us form before e-shoppers could use the company's virtual logo software (which allows them to upload their logo and then see how it would look on a promotional product), Blue Soda increased its overall lead capture by 31 percent and sales by 16 percent.
A/B testing is only part of the data-to-insight-to-action process. Extracting insights from live chats, customer emails, and sales phone calls can be a highly manual process. All of these insights ultimately funnel into Blue Soda Promo's CRM system, which produces weekly, monthly, and quarterly trend reports. To make this "insight dump" as effective and efficient as possible, Blue Soda Promo has identified specific questions that are asked of all customers regardless of which channel they use. For example, customer-facing employees always ask, "What type of event do you need the products for?" This answer, depending on how the conversation develops, can lead to other standard questions related to product purpose, budget, and more.
Passion and process
L.L. Bean has learned how to bring greater rigor to the process through which it derives customer insights from the online ratings and reviews customers provide. Bean implemented Bazaarvoice's Ratings & Review tool in mid-2008 and has since collected more than 300,000 customer reviews. Some of those reviews have boosted sales by restoring previously mothballed products, like the Oxford shirt. Other customer reviews-such as, "I am in love with a doormat. Is it really wrong to love a doormat?"-have appeared in the company's print catalog to advertise products.
An early statistical analysis of ratings and reviews Fuller had requested showed the company that the higher a customer rates a product, the greater that customer's future value is. While that finding passed the common-sense test, Fuller and his team wanted to address an equally important type of customer feedback: negative reviews.
Today, when a product receives six or fewer 1- or 2-star reviews, someone from the product team emails the customer directly to:
1) Apologize for disappointing the customer
2) Suggest an alternative product, if possible
3) Reinforce the company's original guarantee (founder Leon Leonwood Bean said customers could return any product, anytime, for a full refund if they were ever less than 100 percent satisfied), which Fuller asserts represents the "heart and soul" of the company
4) Thank customers for writing the response and, specifically, for helping the company to help them
If a product receives more than six 1- or 2-star reviews, the product team huddles to understand if a more systemic issue exists. Roughly 75 to 80 percent of the time, the poor reviews stem from "copy confusion"-failing to properly describe the product's end use or fit-which can be corrected fairly easily. If that's not the case, the product team decides whether to improve the product (if there is "redeeming value" in doing so) or to remove it from its offerings. If removal is selected, the team then decides whether to liquidate the product, donate it to charity, or destroy the product (because the risk of keeping the company's brand logo on the product is too high).
Last year Bean's customer insights led to the destruction of about $500,000 worth of products, almost all of which were a result of a supplier failing to adhere to the specifications Bean provided. And Fuller says that returning to the supplier armed with the original product specifications and a fistful of disappointing customer reviews represents an extremely convincing set of insights to support its request that the supplier return Bean's money.
The value of that insight-driven action extends well beyond the money reclaimed from supplier errors, Fuller adds, because it helps Bean continue to make good on the guarantee of 100 percent customer satisfaction that has been the lifeblood of the company since Leon Leonwood opened its doors in 1912.