Many years ago as a young Business Unit Executive at IBM, I was assigned an account that was expected to develop into a $50M+ opportunity. Or so my quota said! We had a crack team assigned and everyone was busy, but I wasn’t entirely sure that our relationship was on track. So, I set up a meeting with the CEO to ask: “How are we doing?”
Although he said everything was “fine,” his body language told me a different story. I dug deeper to get the real scoop and, needless to say, we got busy taking action on the CEO’s top issues. Ultimately, the relationship prospered beyond our expectations.
Nothing beats a face-to-face meeting to get customer feedback. Some researchers think that words express only seven percent of human communication. The other 93 percent include how the words are spoken to convey emotion and body language such as gestures, posture, eye movement and facial expression. Emoticons aside, it’s hard to convey a raised eyebrow or shrug in an email.
Obviously, large enterprises with thousands or millions of customers won’t find face-to-face meetings a practical means of customer feedback. So customer surveys are used commonly to collect feedback, often with the help of Enterprise Feedback Management (EFM) systems. But surveys are but one of the six dimensions of feedback you can use to understand the health of your customer relationships.
Ground Control to Major Tom
For space launches, NASA needs a facility to bring mission feedback into a central location for analysis and decision-making. Likewise, Chief Customer Officers need a Voice of Customer Command Center to cope with today’s multi-channel world. In most cases a VoC Command Center can be a virtual resource—short-sleeved white shirts and buzz cuts optional!
In addition to surveys, customers and prospects “speak” explicitly to your business via information captured from web site forms; unstructured text in comments or emails; and audio recordings captured during call center interactions. And that’s not all. Customer behavioral data from CRM and other transactional systems are example of implicit communication that can signal when a customer is likely to defect—a critical bit of feedback these days when retention is a top priority.
Of course, the hot new communications channel is social media, which can also be used as a source of real-time feedback on what customers and market influencers think of your brand. At Gatorade, for example, marketers built a Mission Control Center with six big monitors to display social media data and visualizations from Twitter, blogs and other public sources. Dell, coming a long way from its tone-deaf “Dell Hell days,” launched a similar social media Ground Control center to track conversations and internalize the feedback.
This is the right concept, but monitoring social media separately is just as limiting as only listening to customers via survey resonses. We have enough silos already; it’s time to develop a next-generation Voice of Customer (VoC) Command Center to holistically manage all feedback channels.
Six Feedback Dimensions
Customer feedback comes in many forms. In some cases solicited, such as a survey invitation. But over the past few years unsolicited and often unstructured (non-data) feedback has become more important. As shown in the chart below, a VoC Command Center should handle:
- Structured feedback: Surveys, web experience data and churn signals are mostly data that can be easily processed with conventional information systems.
- Unstructured feedback: In their raw form, text, speech and social media need help from analytic tools to become actionable information.
Let’s review each feedback dimension in a brief tour of the Next-Generation VoC Command Center.
- Dimension 1: Surveys
Customer surveys remain a solid foundation for any VoC programs. Well-designed surveys are sent to known customers to capture key loyalty indicators and dig into specific issues. EFM systems—from companies like Confirmit, MarketTools, Medallia and Satmetrix—deliver the survey invitations, collect the data, identify high priority responses and alert management to the customers needing attention. Some customer service and contact center vendors have integrated customer feedback capability, such as NICE Systems, Parature, RightNow and Verint.
There’s just one catch: Customers won’t always fill out your surveys.
Confirmit has been one of the more aggressive EFM vendors moving to support multi-channel feedback. Over the past couple of years, through acquisitions and new development, the company now supports web, voice, IVR, SMS and native iPhone channels. Furthermore, Confirmit and its many competitors have been working on the feedback management processes, too. Collecting feedback is important, but putting it to use is also critical.
There’s just one catch: Customers won’t always fill out your surveys. After a recent trip I got several survey invitations, but didn’t complete any of them. One hotel launched a survey from the TV after an in-room check out process. Nice idea, but after the first question I noticed 17 more to go. My taxi was waiting so that was that. Other surveys were too difficult to fill out or showed up so late I couldn’t remember much about my experience.
Key Point: You can’t rely on surveys alone to tell you the complete “truth” about the health of your customer relationships.
Dimension 2: Web Activity
My U.S. consumer study in 2009 found that two-thirds of considered purchases started with a search engine. After searching, the most popular first contact was the company’s web site (40%). Do you know how your web visitors like their web experience? Of course, you could pop up a survey invitation using web-centric solutions from ForeSee Results or iPerceptions, or from one of the EFM vendors previously mentioned. But there are other non-survey solutions you might consider.
OpinionLab enables an opt-in approach using a “[+]” widget on any page to invite customers to give input—a kind of online “dynamic comment card system.” Kampyle takes a similar approach, with a twist. CEO Ariel Finkelstein says their feedback widget can be used for lead generation, because over 60 percent of users leave an email address. Companies that immediately followed up on issues found that by “bringing the human touch into the online world,” they could turn complaints into more deals. Now there’s an ROI that’s easy to calculate!
Online banking has become a vital part of customer experience, and for good reason. As a consumer, I like the convenience. The sluggish economy has also helped make the Web a mission critical channel, contends Tealeaf VP of Marketing Geoff Galat. Banks and retailers want to reduce their cost of operations; a store visit or phone call is far more expensive than a web transaction. But if something goes wrong in a private banking transaction, surveys or popup widgets are not the right approach. Instead, Tealeaf offers solutions to record the user’s web session, which an agent can then play back if the customer calls for help. As user experience experts know well, what actually happened may not match what the user says happened!
Key Point: Innovations in web experience feedback can improve the user experience, reduce service costs and even turn leads into deals.
Dimension 3: Churn Signals
Years ago I had a DSL account with an ISP. The service failed, but I didn’t get it fixed for a few months, due to travel and a hectic schedule. But, my credit card kept getting billed right on schedule. Eventually I got an offer from a competing service and decided it was easier to switch than fix. Why didn’t the ISP “connect the dots” and figure out that sooner or later I would probably cancel a service I wasn’t using? A phone call or email might have saved my account.
An often overlooked form of feedback can be found in the transaction and historical data already being captured by CRM and other systems. Customers are speaking to you implicitly via their behavior, captured in the footprints they leave behind in your systems.
Churn models are popular in financial services, retail, telcom—any industry where consumer retention is a top priority to maximize customer lifetime value. As well stated in the American Marketing Association article :
A way to manage customer churn is to predict which customers are most likely to churn and then target incentives to those customers to induce them to stay. This approach enables the firm to focus its efforts on customers who are truly at risk to churn, and it potentially saves money that would be wasted in providing incentives to customers who do not need them.
Customers are speaking to you implicitly via their behavior, captured in the footprints they leave behind in your systems.
For example, in the telecom industry, customers might decide to defect for a number of different reasons, including contract expiration, service quality issues, handset change desired or competitive offers. Using sophisticated statistical techniques, a predictive churn model can be built that relates demographic data, usage patterns, service call frequency and payment data to the likelihood that the consumer will terminate the relationship. In banking, churn models would likely consider changes in account balances and interaction patterns.
Key Point: Predictive models can mine insight from customer data to alert you to satisfaction issues that may lead to customer defection.
Dimension 4: Unstructured Text
Some analysts estimate that unstructured data is 80% or more of enterprise information and is growing much faster. This includes documents, emails, web content, etc. Customers contribute their fair share to this tsunami of non-data by completing comment fields in surveys, emailing complaints or filling out web site forms. In the call center, transcribed call recordings and agent notes are other potential sources of unstructured customer feedback.
Until the past few years, it was too difficult to analyze a large amount of text feedback. Recent advances in packaged text mining solutions help business managers uncover whether written text has a positive or negative tone (sentiment analysis) and identify specific issues that caused the consumer to vent (categorization). Text mining is not perfect (for example, sarcasm is hard to decipher) but with proper set up may yield 70 to 80 percent accuracy. Given the choice between fuzzy insight and no insight, most business leaders will choose the former.
In 2007, I wrote about how Intuit uses text mining from Clarabridge to mine written comments in survey responses from TurboTax users. Then in 2009, in Can You Hear Me Now? Top Five Voice of Customer Pitfalls, I profiled JetBlue’s use of text analytics from Attensity to analyze 45,000 comments per month.
Key Point: Unstructured text feedback is a rich source of insight, but requires specialized solutions to efficiently analyze.
Dimension 5: Speech
Despite the shift to digital channels, the phone remains a preferred interaction method for most consumers based on CustomerThink research. Ironically, so-called Voice of Customer programs haven’t traditionally included the customer’s actual voice—routinely captured in call center recording for many years now. Large call centers record calls for quality management purposes, as input to employee coaching/improvement programs.
But in recent years, speech analytics vendors have emerged to uncover what those calls really mean. Speech analytics is offered as a component solution from many multi-function contact center vendors, including Autonomy, Envision, NICE Systems and Verint. Or you can buy from specialty speech analytics vendors like Nexidia and UTOPY.
While methods vary between vendors, the core idea is the same: figure out the meaning of spoken words captured in audio recordings. For example, as I discussed in my article Use Speech Analytics to Reduce Calls That Frustrate Customers and Hurt Productivity, Blue Cross of Northeastern Pennsylvania used Nexidia’s speech analytics to learn that the root cause of excessive call volumes was system problems and poorly designed processes. As Stephen Covey said, “A problem understood is half solved.”
Key point: Speech analytics can help you efficiently listen and understand literally what customers are already saying.
Dimension 6: Social Media
My recent travels include the usual assortment of good, bad and ugly experiences. Like many consumers, I use social media to give praise for surprisingly good service, or to express disapproval like Stephen Colbert’s Wag of the Finger.
For instance, I used Twitter to compliment a burger joint outside of Boston. But I also wrote a blog post with a video to chastise Caesars Palace for a horrible checkout experience. Boston Burger was clearly listening, because I got a “thanks” tweet back. Caesars Palace gave no public indication that they heard my voice. That’s too bad. Complaints in social media are a gift—they can provide a goldmine of insights that can help a company upgrade service, innovate and improve its competitive position.
The challenge, of course, is dealing with the massive quantity of social data, from blogs, forums, Twitter, Facebook fan pages—any publically available web resource. Coming to the rescue are Social Media Monitoring (SMM) solutions like Alterian SM2, Radian6 (recently acquired by Salesforce.com), Infegy’s Social Radar or Visible Technologies, just to name a few.
Complaints in social media are a gift—they can provide a goldmine of insights that can help a company upgrade service, innovate and improve its competitive position.
After this wave of new entrants, text analytics vendors have responded. Attensity’s Analyze solution handles social media feedback along with many other forms of unstructured text as previously noted. Likewise, Clarabridge released a new version with native support for Facebook and Twitter, and integration with SMM tools.
Not to be outdone by the text mining pioneers, business analytics heavyweight SAS introduced an on-demand Social Media Analytics solution, then later in the year added a Conversation Center to enable customer engagement on Twitter. RightNow offers an integrated Cloud Monitor solution, while SAP, to this point at least, appears to let its partners build SMM solutions on top of the BusinessObjects Text Analysis application.
In a remarkably short period of time the IT industry has responded with a wide spectrum of SMM solutions to capture, monitor and act on the deluge of social media data. But, in most cases these solutions are treating the social channel as yet another silo of automation.
Key Point: Social Media Monitoring has rapidly become a mainstream solution available from a wide range of specialty, text analytics and multi-function vendors.
Getting the VoC Mission Under Control
It should be clear that there are robust solutions for each of the six feedback dimensions. While each deserves special treatment, the next-generation VoC Command Center must pull them all together and help Chief Customer Officers focus their actions.
A complete VoC Command Center should include technology to:
- Capture feedback data across all channels and sources, both explicit and implicit.
- Analyze feedback to identify breaking trends and customers needing attention
- Manage the top priority issues to a successful resolution
As already discussed, capturing feedback information is handled by a variety of vendors. But my research finds most companies are not “connecting the dots” between channels.
Fortunately, help has arrived from analytics vendors that specialize in multi- or cross-channel analytics. NICE Systems, for example, supports IVR surveys and speech analytics. Recently it also introduced multi-channel analytics to spot trends and do root cause analysis.
ClickFox has been a pioneer in true cross-channel analytics to identify and model the actual path taken on customer interactions. This can help explain why customers start interacting with an IVR or web site and then abandon to call a (more expensive) live agent. Solving cross-channel problems, as discussed in my interview with Sprint’s VP of Customer Experience Jerry Adriano, can save large B2C companies millions of dollars in unnecessary phone calls while improving customer satisfaction.
Still, robust analytics is not enough. Enterprises need to act on feedback, which has been the hallmark of EFM systems. The EFM industry, now over 10 years old, was pioneered by companies like CustomerSat and Satmetrix. In recent years there has been some consolidation—MarketTools acquired CustomerSat and Vovici was created from a number of different solutions. EFM vendors have also added support for new social and digital channels to varying degrees. But otherwise, EFM players have shown little interest in a more aggressive M&A strategy to offer a full-featured VoC Command Center as described here. Part of this is due to the sorry state of the economy the past couple of years, which has made finding capital too difficult or expensive.
Industry Consolidation Heating Up
In late 2009, IBM entered the market with a Voice of Customer Analytics solution—a managed service that analyzes structured and unstructured data using IBM’s portfolio of developed and acquired analytics software. With some work a large enterprise could contract with IBM to build a VoC Command Center that incorporated EFM solutions, speech analytics or other sources.
More recently, Attensity upped the ante by announcing the Attensity VoC Command Center designed to enable real-time brand monitoring across multiple online and social media sources. If this sounds like Social Media Monitoring on steroids, it is. Attensity has deeper text analytics function than SMM vendors and has actually built its own database of social data. This should prove attractive to large consumer brands where social is in the spotlight. But, integration with EFM and other vendors would still be required to support other channels and management processes.
In the end, the contact center industry may become the home of most VoC Command Centers. Contact center vendors already manage an array of channels and capture speech recordings too. And they have the financial resources required to consolidate the fragmented EFM industry.
…all VoC solutions should come with the disclaimer “some assembly required.”
For example, in May 2011 Verint announced a new Voice of the Customer Analytics Platform to support cross-channel analysis. Note, however, that some forms of feedback must be integrated from EFM vendors, Clarabridge (OEM partner), Tealeaf, etc. Not exactly a one-stop shop, but clearly heading in that direction. In hot-off-the-presses news, Verint just announced that it is
acquiring EFM vendor Vovici, a move that Verint CEO Dan Bodner says “will fill a void in the market by enabling customers to extract tremendous value from this emerging toolset for the chief customer officer.”
Yes, the VoC Command Center is an essential tool for the Chief Customer Officer. Now that Verint has taken the lead in industry consolidation, I believe there’s much more to come—mainly from large players in the contact center or perhaps the Business Intelligence industry. So (just speculating here) you could see Genesys, IBM, NICE, SAS or possibly SAP respond with their own acquisitions.
That said, for the foreseeable future all VoC solutions should come with the disclaimer “some assembly required.” Waiting for industry consolidation to provide a complete solution from one vendor will mean losing an opportunity for leadership.
Going back to the personal example I used to open this article, Chief Customer Officers must get a complete picture of the health of their customer relationships. If your organization has been managing feedback independently by channel, it’s time to start building your next-generation VoC Command Center.
- Customer Experience Management: Past, Present and Future
- Time to Harmonize Your Cross-Channel Customer Experience
- Use Speech Analytics to Reduce Calls That Frustrate Customers and Hurt Productivity
- Blind Spots in Your Online Customer Experience
- Use Text Analytics To Listen to Customers on Their Terms
- Can You Hear Me Now? Top Five Voice of Customer Pitfalls
- Find the “Ultimate” Loyalty Metric to Grow Your Business
Disclosure: This article was prepared through independent research. Selected vendors are mentioned to illustrate specific capabilities and market developments; no endorsement is implied. Vendors were not included based on sponsor or advertiser relationships. Please visit our sponsor page for information on companies that have supported the CustomerThink community in the past year.