About 18 months ago, Charming Shoppes launched a customer insights project to "deliver actionable customer and market research and analysis to the business," according to Jeffrey H. Liss, who headed up the initiative. Liss is now senior vice president of corporate strategy at the plus-size women's clothing retailer.
Before that time, the company collected and disseminated customer feedback in a less than organized way, Liss recalls. Various departments and brand groups received input from customer emails and online product reviews, and store personnel received verbal comments from shoppers. Anything deemed relevant was "passed up the command chain" to top executives via email distribution lists, Liss says. As a result, "We had a lot of anecdotal information floating around," and executives had no way to distinguish important data from rumor, he reports.
In-house or outsourced?
About 90% of software vendor Clarabridge's customers have chosen its SaaS offering, according to CEO Sid Bannerjee.
This is not surprising, given the costs and headaches of deploying and maintaining a VOC infrastructure in-house. Companies need a data warehouse for customer information, says James Kobielus, an analyst at Forrester Research. "Think of Twitter streaming in all the time. Even if you filter for a subset relating to the demographic you're tracking, or your own customer base, it quickly becomes too much."
Another challenge is integrating VOC tools and platforms with a company's legacy CRM and BI systems. "You want to pass feedback data and alerts into your CRM application so when you call up a customer record, you see not only what they just bought, but how they rated their last interaction," says Bruce Temkin, a managing partner at research firm the Temkin Group.
Conversely, CRM systems can help VOC systems prioritize surveys and responses to customers based on which are hot prospects, or big buyers. Most VOC platform vendors offer their own proprietary dashboards and querying tools, and they're starting to link up with leading CRM and BI systems. BI and CRM vendors should enter the VOC market this year, according to Forrester.
But even with a hosted VOC platform, designing and deploying the infrastructure is far from simple, early implementers agree. "The big challenge is integrating unstructured VOC data with structured databases, in particular CRM data," says Jeffrey H. Liss, senior vice president at Charming Shoppes. "Our CRM group is working toward that goal."
After a considerable amount of research and thought, Liss came back with the concept of a "voice of the customer" (VOC) infrastructure that would collect both quantitative and qualitative input from various customer feedback channels, analyze it for sentiment, meaning and importance, and send relevant data to the right people for further analysis and action.
This infrastructure is becoming even more critical as the company adds new feedback channels, such as an online survey tool that will ultimately deliver approximately 10,000 customer comments a week, according to Liss.
"Sentiment analysis is key," when it comes to interpreting such comments, he adds. For example, "if a customer says, 'I really love going to Fashion Bug, but I don't like sorting through all of the jeans to find the ones that fit me well,' you need to parse the statement using sentiment analysis to understand that she is a big fan of Fashion Bug, but we may have a customer service issue to address," he explains
In December 2010, Charming Shoppes signed up for the software-as-a-service version of a VOC system from Reston, Va.-based Clarabridge. Deployment of the system, Clarabridge Enterprise, is very much in the early stages, says Liss, pointing out that "it takes time to learn how to harness the power of this tool."
Charming Shoppes is hardly alone. While plenty of companies are deploying VOC programs, the majority of existing VOC initiatives are still in the beginning stages. A third-quarter 2010 survey by Temkin Group found that, of 105 companies with formal VOC programs, 63% were still "figuring out what to collect, and how," says Bruce Temkin, a managing partner at the Waban, Mass.-based research firm.
But that's not to diminish the importance of VOC programs in general. A fourth-quarter 2010 Forrester survey of 118 customer experience professionals found that 52% had a VOC program in place and 29% were actively considering one. "Big companies have finally embraced the link between customer experience, loyalty and long-term financial success," says Forrester analyst Michael McInnes. "Investing in voice-of-the-customer programs is the next logical step."
Indeed, businesses are recognizing the value of customer input for a growing number of strategic areas, including marketing and core business processes like product engineering and quality assurance. And it isn't based on customer feedback alone; companies also collect comments and criticisms from industry pundits and the general public.
VOC gets social
Also driving VOC programs is the social Web's growing clout as a consumer sounding board. In a first-quarter 2011 consumer survey by Temkin Group, about 20% of the respondents said that they had reported a bad experience on Facebook, and 13% said that they had reported a good experience on the social networking site. Moreover, 11% had reported a bad experience on a third-party review site like Yelp or TripAdvisor, and 7% had used such sites to report good experiences.
If these numbers seem small, consider this: Only 21% of the respondents said they sent feedback directly to a company via a phone call, a letter, an email message, or its website. (Moreover, the use of social networking sites to voice opinions about companies is likely to grow, since the social Web is a relatively new phenomenon.) In contrast, 63% said they complained about a bad experience to friends via email, over the phone or in person -- out of a company's hearing, so to speak.
Customer data shaping policy
Gaylord Entertainment, a hotel chain, is using Clarabridge Enterprise to gather and analyze data from internal sources, such as customer satisfaction surveys, and external sources, such as industry-specific forums and social media sites like TripAdvisor and Yelp, according to Shawn Madden, Gaylord's executive director of operations analysis.
The platform lets Gaylord employees analyze text to determine what sentiment-scored words are most often associated with which product features in customers' comments. For example, they can determine how often the words clean, small, comfortable or quiet are associated with the word room, Madden says. Employees can then analyze the data to find "moments of truth" -- areas of customer experience where improvement is likely to have the most positive impact, he adds.
For example, such analysis revealed that guests who had a good check-in experience were less likely to complain about other things, while those who had a bad check-in experience tended to look for further faults. They also found that the percentage of customers satisfied with the check-in process plunged from 79% if it lasted less than five minutes to 50% if lasted five to 10 minutes, and down from there.
That insight inspired Gaylord to try to speed up the check-in process in an effort to improve customer experience and loyalty, Madden says. The company created check-in time baselines and targets and offered staffers incentives for meeting them. It also figured out a reliable way to measure the process through daily reports. The payback: The percentage of check-in times under five minutes rose from 39% in the fourth quarter of 2009 to 49% in the first quarter of 2011. During the same period, the average customer rating for ease of check-in rose from 2.4 to over 4, on a scale of 1 to 5, Madden reports.
"Our voice-of-the-customer strategy is integral to our achieving our ever-increasing guest satisfaction goals," says David C. Kloeppel, the hotel chain's president.
Still, many business leaders remain wary of using data garnered from social media, which can be less than accurate or reliable, to say the least. Temkin Group's third-quarter 2010 survey found that only 22% of VOC programs were currently using social media sources, although 35% were considering doing so.
Business executives and business analysts want to ensure that the feedback data they incorporate into critical decisions is of comparable quality to the internal data they've been using. And IT executives need to ensure that their staffs and systems aren't overwhelmed by a flood of largely irrelevant and/or low-quality data.
That isn't stopping some companies from incorporating valuable social media data into their VOC programs. But rather than try to "boil the ocean," as one analyst put it, they are limiting their range to sources that are specific to their products and customers. Charming Shoppes, for example, is looking at monitoring its Lane Bryant customer community site, known as Inside Curve, and its Facebook fan pages, Liss says. "Our customers tend to be vocal and active on plus-size women's blogs," he adds.
During the past few years, customer intelligence (CI) professionals such as marketing and brand managers have increasingly turned to social media intelligence services like Radian6, Scoutlabs and Buzzmetrics, which gather customer feedback from the social Web. The services then analyze the data for relevance and sentiment and present the resulting intelligence in prepackaged reports, charts and "social dashboards."
Such services can cast as wide or as fine a net as customers want; they also offer some degree of quality control. Dow Jones Insight, for example, "selects social media based on how influential it is, how frequently it was updated in the last 90 days, and whether it is free from spam and porn," says Martin Murtland, the service's managing director. It can also add targeted sources at a customer's request, such as "Twitter feeds based on client-requested content," he notes.
The problem is that such deployments tend to create information silos that are isolated from IT staff and systems, and as a result there's little sharing of insights among groups. VOC programs need to integrate all of the various feedback channels into a single infrastructure, Temkin points out. And that's where IT comes in, he adds.
VOC tools: It's a fractured market
At this point, "no single vendor provides [the] full functionality to meet the entire social intelligence needs of the enterprise," says Forrester analyst Zach Hofer-Shall. However, many vendors have been working hard this past year or two to expand their platforms to cover all the VOC bases.
But if you're interested in investing in a VOC system, you should do your homework. First of all, you need to know what market the vendor started out in, because that represents the company's core strength. Recently added or acquired features and functions may not be fully integrated with its earlier platforms, warns James Kobielus, an analyst at Forrester. Vendors are aggressively integrating VOC functions with business intelligence, CRM and advanced analytics tools -- and "when you have a laundry list of technologies, a lot of assembly is required," Kobielus says. A systems integrator could be helpful in working with such a vendor, he suggests.
Social intelligence service providers mine the social Web for customer feedback, then use sentiment analysis and natural language processing to determine its relevance and sentiment score. Major players include Radian6, Netbase, BuzzMetrics (now part of Nielsen), Crimson Hexagon, Scout Labs (now Lithium Technologies), Cymfony, NM Incite, WiseWindow, MediaMiser, EmPower Research, Synthesio, Converseon and Dow Jones Insight.
[Editor's note: To see a page with links to all of the vendors mentioned in this sidebar, click here.]
Customer insight and action platform vendors have strong backgrounds in mining and analyzing large bodies of unstructured material generated by internal channels such as customer surveys and call center notes. Major players in this field include Clarabridge, Attensity, Kana, Autonomy, SAS Institute and RightNow. Many of these companies have recently moved into the social media market by partnering with or acquiring social intelligence service providers. Their suites often include social community software, as well as BI and CRM tools -- either their own or those of a partner.
Social community platform vendors such as Lithium, Jive Software, Telligent Systems, Kickapps and Mzinga help businesses set up online communities where customers can chat, exchange tips and express opinions. They also provide tools for monitoring, mining and analyzing customer interchanges. Some, like Telligent and Lithium, allow businesses to set up their Facebook fan pages so that a customer who posts a question there automatically gains access to the social community, which often has the answer.
Enterprise feedback management (EFM) vendors like Allegiance (now part of Nielsen), Mindshare, Medallia, and ResponseTek specialize in tracking and analyzing customer feedback and behavior, and the effectiveness of media campaigns and other marketing strategies.
At iRobot, for example, customer feedback used to reside in a variety of information silos, including outsourced call centers and a growing number of social media sources, says Maryellen Abreu, director of global technical support at the maker of self-guided vacuum cleaners and other robotic equipment.
This meant that managers had trouble correlating the data to make high-level decisions. For instance, engineering didn't have immediate visibility to data in order to make rapid decisions on changes in product design, Abreu says. "We're introducing new products all the time, so it's important to have that immediate, almost-real-time feedback," she adds.
Burlington, Mass.-based iRobot chose to use RightNow CX, a SaaS-based customer experience management system from RightNow Technologies that "gives a 360-degree view of the customer: when they called, emailed, chatted or posted on a forum, and what issues they brought up," according to Abreu. It aggregates opinions from various sources so that top managers can quickly determine customers' key concerns and then respond in a timely fashion.
"Timely" is a relative term, however, when it comes to the social Web. "Because our product is very visual, customers would show us problems on YouTube, and at the end say, 'iRobot, what are you going to do with this?' and the post would have 60,000 hits," says Abreu. "We needed to respond faster."
The robot maker now uses RightNow's Cloud Monitor, which mines customer posts for words or phrases with negative connotations, such as all-cap words and "bad language," Abreu says. It then alerts customer support personnel and, if a post starts to go viral, automatically escalates alerts to Abreu's attention.
Less talk, more action
Right now, most companies are still figuring out the critical components of their VOC programs: what data to look for, what metrics to use, and most important, what action to take, according to Temkin. "Feedback is cheap, actionable insights are priceless," he says.
IT and business leaders should not become discouraged: Even early stage VOC programs can get good results, according to Temkin. "Once you get into actually quantifying how customers view you, it starts changing how your people think about the business," he explains. "They start to spot customer issues and put in place processes where they can highlight and start to solve the big problems. And the big payback is customer loyalty."
Pricing for some VOC products
Lithium (social community, SaaS): The full suite, for midsize to large companies, costs $7,500 to $10,000 per month.
Dow Jones Insight (social intelligence): $10,000 to $100,000 or more per month, depending on the size of the installation, the number of topics customers want analyzed and the number of languages supported.
Clarabridge Enterprise (customer insight/action): Pricing starts at six figures per year and depends on volume, services and support.
WiseWindow (social intelligence): $10,000 to $12,000/month
Liss definitely has some ideas. For example, monitoring blogs will give his employees an early heads up on buying trends. "If all of a sudden plus-size women are talking about how comfortable a certain fabric is, we can study it for use it in our products," he says.
Charming Shoppes' VOC team has already developed a dashboard, including a one-page document that each brand group uses in its monthly business review. This enables groups to share their insights, according Liss. "Right now it's just the highlights -- major feedback that's actionable -- but we'll build from there," he says.
Sign up for Computerworld eNewsletters.