In the first of an occasional series of guest posts on the Customer Technology blog, we're lucky to hand over to Kurt Marko - technology writer, independent IT consultant and editor-at-large for InformationWeek and Network Computing.
The nexus of mobile, social, cloud and big data is radically reshaping the business world, nowhere is the upheaval more dramatic than in areas with high customer engagement: retailing, financial and communications services, travel, customer service and technical support. Such profound changes in the landscape, driving heightened consumer expectations have pushed beyond the consumerisation of technology to the consumerisation of customer experience. The effects show up in everything from client devices and applications to IT infrastructure and application provisioning. It's permanently altered consumer behaviour, notably in how people learn about, evaluate and buy things, how they share opinions about products, retailers and service providers with an ever-widening circle of online friends and the ways they expect to access documentation and support. People routinely learn about a product while checking Facebook or Twitter from their smartphone, research a purchase using a tablet, compare alternatives at a retail showroom, buy it using a mobile app and call for customer support. These interactions, which span multiple platforms and communication channels in both the physical and online worlds are what we call the omnichannel customer experience, and it is both a reaction to and driver of heightened consumer expectations.
Omnichannel is best known as a marketing and sales strategy popularised by retailers like Apple, Macy's, Nordstrom and Target that seek to provide a seamless buying experience between online and in-store shopping. With online shopping growing at double-digit rates, one survey found that two-fifths of shoppers will make the majority of 2013’s holiday purchases online [PDF]. Indeed, estimates for the 2013 holiday shopping season predicted that mobile-optimised retailers would generate more than a fifth of their online sales from mobile devices, while overall about 14% of online sales came from mobile customers, an increase of 40% since last year on busy shopping days like Black Friday.
But the omnichannel imperative isn't just limited to retail, nor does it end with the purchase of a product or service. Customers expect a consistent, smooth and unified experience in all their interactions from buying and purchase to customer service and technical support. This “customer journey” includes the steps that comprise the customer experience lifecycle: browsing, researching, evaluating, buying, using, sharing, supporting, repairing and upgrading. Stitching the seams between these steps and the different interaction channels used during a customer journey is challenging. The advent of inexpensive storage and sophisticated data analytics paired with predictive algorithms, however, can turn vast quantities of previously unusable data into useful, actionable information. Layer on the power and convenience of cloud services, which can deliver big data analytics at scale without expensive up-front capital costs, and it means omnichannel customer service is no longer just a pipe dream. In this white paper, we'll explain the trends and opportunities, what they mean for business executives and how real time predictive analytics can transform the multiplatform, mobile, social customer experience.
Omnichannel customers and the unified experience journey
The traditional notion of targeting multiple channels for customer engagement and interaction comes from marketing where businesses learned to tailor messages and advertising content to different communications media: print, TV, direct mail flyers and catalogs, email and more recently, mobile devices. However, these were often distinct marketing campaigns with their own messaging and rarely tied to company-wide CRM or product support systems. Translated: multichannel often meant siloed experiences focused on sales that targeted a specific medium, whether online/Web, phone or in person. A senior executive at a major cable company says, "Multichannel is an operational view –- how you allow the customer to complete transactions in each channel." The siloed approach worked well enough when customers started and finished a transaction or interaction in a single channel, say calling customer support to resolve a billing discrepancy, researching and buying a product over the Web or visiting a retail store to add a new phone to their wireless plan. In today's mobile, hyper-connected world, however, where we're all afflicted with a form of smartphone-induced ADD, people increasingly multitask, interacting with multiple devices over several channels to complete a task – the web, email, social networks, text messaging, voice -- initiating a conversation or task on one, taking it up later on another.
But the power of omnichannel to transform industries and business models is on full display in everything from the increasingly symbiotic relationship between TV and social networks, particularly Twitter, a phenomenon known as social TV, to how airlines interact with their customers using websites to book flights, mobile apps for boarding passes and text messaging for status updates. These platforms, from social networks to mobile apps and instant messaging, provide new real-time communication channels between users and service providers. Whether its a customer praising the quality of a new specialty drink on the Starbucks Facebook page, getting product support from product-specific discussion groups, or taking a product satisfaction survey via a series of text messages, there are now dozens of ways businesses can engage and interact with their customers.
Maximising the value and effectiveness of these various connection points requires making sure the customer experience delivered across them is integrated and not disjointed. In essence, a business needs to provide its customers one digital channel, spanning multiple devices and screens that provides a consistent and contextual customer experience; one that understands where a particular activity left off so that it can be resumed at another time or on another device. Furthermore, it should be predictive; good at anticipating what the customer wants to do next to complete a particular task. For example, when someone calls your help desk, your agents should know if the customer has been searching your online knowledge base about a particular problem. This is the essence of the omnichannel customer experience: it's seamless and smart. As the aforementioned cable executive put it, "Omnichannel, however, is viewing the experience through the eyes of your customer, orchestrating the customer experience across all channels so that it is seamless, integrated and consistent. Omnichannel anticipates that customers may start in one channel and move to another as they progress to a resolution. Making these complex 'hand-offs' between channels must be fluid for the customer. Simply put, Omnichannel is Multichannel done right!"
Customer experience challenges in an omnichannel world
Today's mobile, multi-platform reality can end up creating sub-optimal customer experiences in a number of ways. Some of the most common problems (if you're a customer) or challenges (if you're a business) include:
● Inconsistent information: As customers move from channel to channel, they often encounter inconsistent information, presented using widely different designs with no easy way of synchronising content between journeys along different channels. While discussing shopping trends, Neiman Marcus' President and CEO Karen Katz noted that the new generation of customer expects to shop between channels, citing a customer survey finding that 70% start their shopping online. She said most "start their journey online, doing a lot of research online to understand what the landscape looks like and then either buy online or come into the store." Not only is having independent platforms for Web, iPad, iPhone and Android extremely inefficient, it makes for a confusing customer experience when some actions, say booking a reservation or initiating a customer service request, are available on one, but not the other.
Many customers can surely relate to Erin Levzow, marketing director at Palm Casino Resort when she says, "My least favorite thing is when you go to a website and it offers the user to enter the full site. That means that not everything is included on the mobile site." Instead, Palm redesigned its online presence to embrace omnichannel. Says Levzow, "We've taken the mobile site experience and want to make one strong experience. What's so amazing about the platform that we use is that if you make a change it goes live on all devices." However, pulling this off requires more than a redesigned mobile app, but a comprehensive, omnichannel-aware mobile strategy.
● No personalisation: Having been conditioned by years of using online social and ecommerce shopping sites, consumers expect their digital experience to be tailored to them: to record their preferences, learn their habits and keep track of them regardless of the platform they happen to be using. As Thomas Davenport, distinguished professor at Babson College, illustrates in recounting his own experience in using a myriad of services, consumers facing "a great deal of potential privacy lost, but dramatically-increased convenience and value" generally opt for service and convenience. "It’s a tradeoff I am willing to make," writes Davenport.
But omnichannel personalisation means more than just synchronising wish lists between mobile apps and a Web site or offering purchase suggestions based on what's in your current shopping cart. In today's mobile-centric world it requires recording when and where a customer performs a particular action and using present context and past history to improve the customer experience in creative ways, for example by warning a wireless phone customer about running up against their data cap because more of their activity is on LTE versus Wi-Fi while traveling.
● Frustrating customer support experiences: Everyone has suffered through a product support call where you transfer two or three times and each agent asks for the same routine information. Omnichannel can amplify the frustration through the sheer increase in the potential number of handoffs. A customer may initiate contact via a mobile app, come back later to fill in problem details using a Web form and finally call a live agent for a one-on-one consultation. Asking them to answer the same questions two or three times increases customer effort and frustration, which eventually leads to increased customer churn and attrition. In contrast, good customer experience increases loyalty, which a Forrester study finds not only reduces costs, but leads to increased revenue. Happy customers are loyal customers and loyal customers are buying customers.
● Incomplete, inconsistent customer intelligence: Poor tracking of customers across channels doesn't just harm them, for businesses it means losing potentially valuable information about customer wants, needs and behavior. Without integrated omnichannel data and analysis, companies forfeit growing opportunities to enhance the customer experience by providing more relevant content, more effective self-service tools, cross-selling recommendations, accurate and timely alerts and better loyalty programs.
● No sense of community, nor leverage of social networks: Social is such an indispensable element of today's online existence that regardless of the product or service, it can no longer be treated as an isolated feature, but must be woven into the entire customer experience. Whether through sites integrating with public networks like Facebook, Twitter and LinkedIn or captive commenting and review systems like those popularised by Amazon, people expect the customer experience to be a two-way conversation. A disjointed omnichannel experience, particularly one without a compelling mobile presence, makes building a community of loyal, engaged customers a virtual impossibility when you consider the average consumer uses three or four devices and platforms a day, stealing time wherever and whenever they can to be productive and get things done.
Predictive Analysis: Key to Seamless Omnichannel Experience
Addressing these omnichannel challenges requires more than just coating over your website and mobile apps with the same UI design, it requires a multi-pronged, data-management strategy in which content and decision support applications all tap into the same, consolidated repository. Regardless of the absolute size of your data set, this is a big data problem since it requires capturing and analysing 100 percent of your customer interaction data; everything from web transactions and online chat logs to CRM records and contact center transcripts. Yet data without organisation and analysis that enables action is just a scrap pile. Furthermore, in today's world, where the time value of data rapidly decays, analysis that takes hours, not seconds, is often useless. Thus, an engaging, effective omnichannel rests on three pillars: big data, predictive analytics and real time engagement.
Predictive Analytics 101
Predictive analytics is one of those concepts easily understood in the theory, but deceptively complex to implement in practice. Still, predictive software is something we all rely upon many times a day. The basic idea is using data about what's happened in the past, along with current context, to predict what's likely to happen in the future.
The most common and well developed example is weather forecasting, which uses a vast set of temperature, wind, barometric pressure and other atmospheric data collected at thousands of spots around to globe to feed sophisticated mathematical models of weather conditions. Every few hours, new data is collected, the model is initialised to current conditions and run forward in time, producing what is often called a futurecast. The quality of the forecast depends upon several factors: the precision and volume of data, the accuracy of the underlying scientific model, the spatial and temporal granularity of the simulation (including more data points in both space and time improves the forecast) and the model's ability to self-correct or improve over time as it compares past predictions to current results.
These same attributes affect the accuracy of software designed to predict and correlate customer behavior. An omnichannel customer engagement platform is only as good as the data it has to work with, thus the more comprehensive a company's data collection strategy, the better and more actionable the analysis. Likewise, the more accurate and timely the data, the more effective predictive analytics can be at anticipating customer needs and behavior. Garbage in, garbage out still applies.
Much like the weather, customer behavior and demographics are constantly changing, so omnichannel predictive analysis must quickly and accurately learn from the past. For example, if a financial firm's fraud detection software flags a customer's credit card charge at Tiffany's on Christmas Eve because it's the first time a customer has done business there and a larger transaction than normal, once an agent verifies that the charge is legitimate, the software should be smart enough to learn that, a couple months later before Valentine's Day, a similar-sized charge at Cartier's in the same mall isn't unusual. But as the fraud detection example illustrates, predictive analytics must run in real time. There's no point in learning a transaction is fraudulent after you've already sent the bill and triggered an angry customer complaint.
In the context of omnichannel, an example would be that a large percentage of customers are purchasing from a website. A high proportion of customers that make a travel purchase over $3,000 in value tend to call customer service immediately after placing an order on the web to ensure the transaction has gone through. Predictive analytics in omnichannel can potentially provide a proactive chat notification or click to call at the end of the transaction to get that person directly in front of a live agent – seamlessly and with less effort and time from the customer.
Finally, not only must omnichannel analytics ingest and use data from all contact points -- mobile, Web, phone, etc. -- it should also support customer interaction using the same set of channels. If a customer initiates a support question via a mobile text message, they sure don't want nor expect an email response.
Gone are the days when a company can steer customers to the channels it prefers and controls. Consumers today use the channels and devices they consider most convenient and effective for the task. This means companies need a comprehensive data acquisition strategy that incorporates the key channels customers want to use. Yet delivering a stellar omnichannel experience also requires sophisticated software that can analyse transactions, track customer journeys across sessions and channels and work with multiple output channels; software that's beyond the scope of what most companies can build on their own and the result of significant R&D efforts and intellectual property.
Cloud economics and convenience, as demonstrated by Google's AdWords product, itself a highly scalable and distributed predictive analytics platform, mean that omnichannel platforms are typically best delivered via a SaaS product. However this doesn't mean a company's sensitive customer data must reside in the cloud. Out of concern over regulatory compliance, customer data security and company intellectual property, many IT departments want the comfort of managing their own data and don't want to redesign existing data collection processes, which may pull from chat and voice logs in the contact center, CRM systems, product registration databases, etc., that feed their big data repository. Thus, SaaS products should offer some form of hybrid model in which bulk data is stored locally and securely packaged and transported for analysis and decisioning.
In sum, delivering a leading omnichannel customer experience requires: comprehensive data collection across channels (big data), scalable predictive analytics software with adaptive algorithms that manages customer journeys across channels, support for all common communication channels (mobile, Web, voice, text), real time decisioning and hybrid cloud delivery with on-premise data repositories feeding cloud-based SaaS applications.
For those not yet convinced that omnichannel is a competitive imperative, look again at trends in the brutally competitive retail industry. While overall sales were off over the recent
Thanksgiving holiday, mobile sales spiked by 40% or more, but it's not all just going to Amazon and eBay. According to Vicki Cantrell, SVP of the National Retail Federation, savvy traditional retailers are merging online and on-premise shopping to optimise both their marketing clout and inventory management. “Retail is becoming increasingly digital, which gives retailers the opportunity to utilise their stores in support of this and enhance the entire experience for the customer,” she says. Forrester analyst Sucharita Mulpuru reiterates the point saying, “Retailers need to be wherever customers are, which these days means online and on their phones.”
But retail is just the canary in the omnichannel coal mine as the phenomenon of mobile-first consumers expecting to seamlessly wander between multiple communication channels will ultimately affect every business. As Forrester's Ron Rogowksi blogs, "companies need to focus on delivering great experiences, day in and day out, with products and services that meet customer needs, are easy to use, and are enjoyable." Are you ready?
Kurt Marko is a technology writer, an independent IT consultant and editor-at-large for InformationWeek and Network Computing. His career has spanned virtually the entire high-tech food chain from chips to systems including positions at AT&T Bell Laboratories and Hewlett Packard. Kurt has been a frequent contributor to several IT trade and consumer technology publications and industry conferences. He holds a BS and MS in Electrical Engineering from Stanford University.