It Ain t Personal; Get Up Close And Contextual

(1)

Contextual

by stephen powers, Anjali yakkundi, and Ron Rogowski, november 19, 2012

& Delivery professionals

key TakeaWays

Firms Will Maximize Customer Relevance With Contextualization Firms must deliver relevant experiences to their customers across channels. But fi rms have struggled to do so from both a strategy and a technology standpoint. Th e answer is to deliver more-contextual experiences that are tailored, adaptive, and

sometimes predictive and support those experiences with multiple technologies. decide Which processes Will support Contextualization initiatives Two main processes can support contextualization initiatives: rules-based manual

processes and automatic processes. Rules-based manual processes require creating rules that dictate when content is displayed. With automatic processes, delivery applications display based on previous behaviors and/or expert-driven metadata. Map Web Contextualization needs To existing Tools

Th ough the vendor landscape for products that support contextualization is fragmented and confusing, you may already own at least some of the technologies needed to support contextualization. Th ese include products such as web content management (WCM), site search, analytics, eCommerce, recommendations engines, and analytics solutions.

(2)

Why Read This RepoRT

Organizations have long “personalized” their websites for customer segments. Now they need to go further, first by delivering the right experience to the right user at the right time and on the right device and next by delivering relevant, tailored experiences that meet individual user needs by combining historical, behavioral, and profile data with real-time situational feedback. We call this next step in customer experience targeting contextualization. This report helps application development and delivery (AD&D) professionals understand what contextualization is, what its benefits are, and what tools, processes, and technologies can help deliver it.

notes & Resources

Forrester interviewed 16 vendor and user companies, including: 3m; Adobe; Autonomy, a division of Hp; Baynote; Bluefly; coveo solutions; Dell; Ektron; isite; monetate; Oracle; Razorfish; RichRelevance; sDl; siteworx; and UncommonGoods.

supporting Relevant Digital Experiences In The multichannel World

by stephen powers, Anjali yakkundi, and Ron Rogowski

with Harley manning, John R. Rymer, Allison stone, and Rowan curran

2 3 5 7 8 9

(3)

Today’s CUsToMeRs deMand ReLevanT expeRienCes

More than ever, firms need to deliver relevant experiences to their customers across a range of devices.1_{These experiences must:}

■

Meet customer needs. A highly relevant experience meets the content and functionality needs

of a customer while satisfying the underlying emotional needs that trigger an interaction.2

■

Feel personal. Consumers expect — and appreciate — personalized experiences.3_Truly

personal experiences deliver relevant content and function based on explicit and implicit feedback about customer needs and preferences.

■

Deliver in the moment. Relevant experiences take into account a person’s current state — such

as time of day and location — and deliver on situational needs, such as finding additional product information using a bar code scanner at the point of consideration.4

Firms struggle To deliver Relevance at scale

What stops firms from delivering highly relevant experiences? Companies fall short of meeting customer expectations when they:

■

Guess at what their customers want and need. Despite the proliferation of tools such as

personas and journey maps, in a recent survey, only 41% of the 86 CX professionals from our ongoing Marketing & Strategy Research Panel said that they consistently use customer research as an input into customer experience design projects.5

■

Don’t share customer data across silos. Nothing says “I don’t care about you” more than

forgetting a customer from one moment to the next. But that’s exactly what companies do when they fail to share customer information across organizational or channel silos.6_{As a result, when}

customers cross from one touchpoint to another — for example, moving from a mobile app to a desktop website — firms fail to recognize them, so customers have to start pursuing their goals from scratch.

■

Don’t take advantage of real-time cues. Users’ current circumstances, such as location, time

of day, and even weather, can determine their immediate needs. But even though today’s connected devices can provide a wealth of information about users’ context, few firms take advantage of this data to predict what users want and need from them in the moment.7

(4)

To MaxiMize ReLevanCe, Think ConTexTUaLizaTion

To succeed in today’s digital environment, firms must deliver smarter, more-customer-centric interactions that feel like they were tailored for each user and his or her specific set of

circumstances. That’s why firms need to evolve their thinking to focus on contextualization, which Forrester defines as:

A tailored, adaptive, and sometimes predictive digital customer experience.

Contextualization combines and extends existing segmentation and personalization techniques with in-the-moment details. This enables more-dynamic, more-predictive experiences by processing explicit and implicit user information.

Three Types of data inputs power Contextualization

Three key types of data inputs, which occur at both the aggregate and the individual level, power contextualization (see Figure 1):

■

Profile: Who the customer is. Aggregate information about the audience segment members

and their behaviors — for example, how they interact with digital channels, key concerns, and subject-matter knowledge — helps companies design the right overall experience. Personal information about individuals, such as where they live and their specific interests, allows additional tailoring.

■

History: What the customer did in the past. People leave many markers when they interact

with a company, including a record of the web pages they visited and the purchases they made. A firm can combine aggregate data from all of its users with individual histories to anticipate

customer questions and/or predict future actions.

■

Situation: What’s happening with the customer now. Factors such as time of day, geographic

location, device, and browser indicate a user’s current situation. Organizations can map this data to aggregate information about similar users to help predict what an individual may be trying to achieve at a given point in time.

(5)

Figure 1 Contextualization Combines Segment, Personal, And Situational Data

Source: Forrester Research, Inc.

80681 History Profile Situation Persona Sale Buy this

(6)

ConTexTUaLized ReLevanCe is poWeRFUL

Contextual interactions manifest themselves in a number of ways, including:

■

Location-targeted content. Location-targeted content can take a number of forms. For

example, GateGuru, a mobile app that helps travelers find restaurants and services at airports, suggests a list of local airports to view based on user location. Home Depot adjusts some of its featured content based on seasonal changes, featuring patio furniture to people in the South in early spring while pitching kitchen remodel products to people in the still-cold Northeast.8_And

to help bolster its mortgage loan business, Commonwealth Bank of Australia created an app that uses augmented reality to provide listings with prices, photos, and data on comparable homes.9

■

Adjacent content. Users’ path through a site or application gives telltale signs of what they’re interested in. Firms can combine this information with aggregate data to deliver related content specific to the browsing patterns of their users. For example, photo equipment retailer Jessops tracks movements and provides dynamic recommendations based on business rules that limit the price differential between the product initially viewed and the recommended product.

■

Right-sized content and functionality. Users don’t want every piece of content everywhere —

they want the right pieces of content and functionality sized for their needs. For example, Amazon. com’s mobile site does not provide the same level of detail as its website, nor does it provide editorial reviews or author information for books. Instead, the mobile site focuses on price and limits product details to the essentials needed to make a decision.

■

Adaptive designs. When a firm needs to support users across devices with highly varied display

characteristics, it can use a liquid layout that will render cleanly — but differently — from one device to another. Sites such as Purina’s catchow.com and Boston.com provide a single set of imagery, navigation, and content that adjusts to the size of the user’s browser, regardless of device.

differing Contextualization Methods process inputs into interactions

When AD&D professionals support contextualization efforts, they must think about: 1) the contextual data points coming in; 2) how that data will be processed; and 3) what output the customer will see.

Two different methods for supporting processes utilize profile, historical, and situational data in order to create contextualized experiences. They are:

(7)

■

Manual rules-based processes. Manual rules-based contextualization allows business users to explicitly define various consumer segments based on input variables. For example, a web page might have an embedded rule that essentially states that if visitor is part of segment X, then the site should display Y. Depending on scope and complexity, purely rules-based initiatives can require a great deal of manual effort, as business users must manually maintain rules that determine which visitors see which type of content. While these methods easily support traditional personalization initiatives (based on profile and historical data), they are often not dynamic enough to support full-blown contextualization initiatives (which include more situational data).

■

Automated processes. Automated processes display content or offers based on automated

filtering of numerous inputs that may be both explicit (e.g., previous purchases and stated preferences) and implicit (e.g., click path, referring site, and search terms). Solutions that support automated processes have built-in algorithms that are self-learning in order to provide dynamic recommendations. Two inputs drive most of the automation: expert and crowd. Expert-driven automation relies on metadata to guide automation (e.g., if visitor A clicks on a washing machine, content related to “washing” — such as dryers — will be recommended to the visitor). Crowd-driven automation relies on “the wisdom of the crowd” and bases results on the collective behavior of users (done through the solution’s self-correcting processes). This process is important for product recommendations and “more like this” content. Automated processes require less effort than explicit rules, but the “black box” approach may not be attractive to those who wish to more tightly control some content targeting.

In addition, organizations implementing contextualization must also determine how they will handle:

■

Segmentation. Segmented personalization processes deliver contextual content to specified

consumer segments using both manual rules and automated approaches. Companies may use broad segments, smaller microsegments, or even one-to-one contextualization, which delivers targeted and unique content to each individual consumer and realistically can only be achieved using automated solutions as opposed to rules. Some technologies even offer segmentation detection.

■

Push versus pull. Companies “push” content to consumers based on profile, historical, and

situational data. But consumers may also “pull” relevant content by indicating explicit, persistent content preferences (such as a preference for football content on a website such as ESPN.com). Companies then store this information as part of the customer profile and use it to deliver content in an authenticated situation.

(8)

■

Cross-channel contextualization. Many organizations still view different channels in silos. The move to contextualization will require that organizations break these barriers and offer cross-channel contextualization. However, outside of retail, many organizations plan to start contextualization initiatives in the mobile and web channels.

The Web ConTexTUaLizaTion TeChnoLoGy LandsCape

Many organizations’ contextualization strategies are relatively nascent. You’ll likely be asked first about web contextualization before other customer touchpoints.

Firms beginning or expanding support for web contextualization may already own at least some of the technologies needed to support contextualization. Both manual rules-based and automated contextualization solutions that draw on profile, historical, and situational data are available in a variety of products that support customer experiences. In particular, organizations will find these technologies available or embedded in:

■

Web content management (WCM) and commerce solutions. WCM and commerce solutions

offer business-user-friendly interfaces that enable manual rules-based personalization.10

In addition, they increasingly offer automated solutions for deeper contextualization via integration with other products in the vendor portfolio (e.g., Adobe CQ5 and SDL Tridion). Some also offer some basic tools for managing segmentation.

■

Behavior-based analytics solutions. These solutions take an automated approach,

contextualizing and targeting content based on environmental and individual variables that provide comprehensive views of customer behavior. Some solutions focus on predictive analytics (e.g., IBM, Oracle, SAP, SAS, and Tibco Software), while others focus on testing analytics (e.g., Adobe).

■

Environment-based analytics solutions. These solutions (e.g., ExactTarget) contextualize

based on environmental factors such as IP address, industry, country of origin, device type and optimal screen resolution, and operating system. These solutions are generally manual rules-based options (though some are moving to a more automated approach, especially with regard to responsive design for device), and they work with delivery systems’ WCM or eCommerce in order to deliver the targeted content.

■

Independent recommendation engines. Third-party recommendation engines (e.g., Baynote,

Monetate, RichRelevance, and SDL Fredhopper) dynamically serve recommended products as well as text, ads, and other content. These solutions primarily focus on commerce sites, though some vendors have expanded outside of this one use case (e.g., self-service or informational websites). Many recommendations engines use a combination of a manual rules-based method and an automated approach driven by crowd behavior.

(9)

■

Site search solutions. Site search products (e.g., Adobe and Oracle Endeca) use automated approaches to contextualization. These solutions can “boost or bury” search result records based on individuals’ previous search queries and crowd behaviors. More-advanced site search can assemble entire web pages and utilize search analytics to influence and contextualize promotions and microsites.

R e c o m m e n d at i o n s

appLy The RiGhT TeChnoLoGies To sUppoRT ConTexTUaLizaTion eFFoRTs

Contextualization can be powerful for companies that know how to apply it. This includes applying the appropriate technologies in the right scenarios. To support contextualization, application development and delivery professionals should:

■

Assess current technology capabilities before purchasing new solutions. Don’t be overly

influenced by glossy and bold promises from new vendors. Instead, evaluate current capabilities included in your digital experience technologies against business requirements before immediately going on a buying spree. You will almost certainly be able to support simple, manual rules-based contextualization with your current tools, and as you progress, you can look to source newer, more specialized solutions.

■

Prioritize marketing self-service. Contextualization solutions truly straddle the line

between BT and marketing departments. BT should stay out of critical publishing paths, instead empowering marketing departments wherever possible. This means that, whether you are ready to source new technologies to support contextualization or you are using existing tools, these solutions should have intuitive user interfaces (UIs) that allow business users to easily create and define segments and manual rules.

■

Create bulletproof, high-octane metadata and taxonomy structures. Many of the

processes outlined in this report (e.g., manual rules-based solutions, expert-driven automation, and known user contextualization) rely on well-defined taxonomies that are tagged appropriately. Without these, contextualization solutions, no matter how impressive the functionality, are likely to fail. Consider issues such as: appropriate taxonomy structures, skills and staffing (this is particularly important, as marketing departments don’t often have a librarian within the department), governance (e.g., how iterative taxonomy improvements will be made), and how business users will be educated on taxonomies in order to ensure that they are properly maintaining and tagging content.11

■

Understand that failure may precede success. As a global director of marketing and

(10)

fifth) attempts may not yield desired results. Contextualization is a largely iterative process, so be patient, take many small steps, and learn from your mistakes as you go. You can start by introducing a couple of segments or by installing a plug-and-play recommendations engine and iterating as new issues come up. In the longer term, you can deploy more-complex contextualization strategies that use a hybrid approach incorporating manual rules as well as automation.

■

Start with a manual rules-based or an automated approach but work toward a hybrid

approach. Many organizations dive head first into a manual rules-based or an automated algorithmic-based approach. But there are pitfalls for each of these. Approaches based too heavily on rules are manual and become tedious for marketing professionals. Purely automated approaches make many companies nervous because targeting is done in a “black box” with little ability to control any of the recommendations (think of the Harry Potter problem — where Harry Potter is recommended to everyone regardless of purchase because of its popularity, which happens when rules are done in the “black box”). While organizations nascent in contextualization strategies can start with either one of the other, they will likely settle on a hybrid approach. As you move toward a hybrid approach, you can also begin branching into cross-channel personalization initiatives to foster more-contextual experiences.

sUppLeMenTaL MaTeRiaL

Companies interviewed For This Report

3M Adobe Autonomy (a division of HP) Baynote Bluefly Coveo Solutions Dell Ektron iSite Monetate Oracle Razorfish RichRelevance SDL Siteworx UncommonGoods

(11)

endnoTes

1 _{Today’s empowered customers own multiple connected devices and have high expectations for getting}

information and services when and where they need them. As of Q3 2011, the average US consumer owned two connected devices. Given forecasts for continued smartphone and tablet adoption, that number will likely increase in the coming years. Source: North American Technographics® Online Benchmark Survey, Q3 2011 (US, Canada).

2 _{Customers can’t always articulate exactly what they want. But rather than merely designing for stated}

needs, firms need to stay one step ahead of their customers by walking the line between researcher and customer — and not just for a single project but as a regular practice. See the January 18, 2011, “Mastering Emotional Experience Design: Address Customers’ Real Goals” report.

3 _{Sixty-two percent of consumers find personalized retail websites useful when shopping online, i.e., when}

an online store recommends products based on other items that the consumer may be browsing or buying (e.g., “You might like . . .”). Source: North American Technographics® Retail Online Benchmark Recontact Survey, Q2 2010 (US).

4 _{Successful mobile services must be immediate, simple, and contextual, which requires using more of the}

available contextual information to deliver more meaningful content and functionality to users. See the October 26, 2011, “The Future Of Mobile Experiences Is Context” report.

5 _{Source: Q4 2011 Global Customer Experience Peer Research Panel Online Survey.}

6 _{Fifty-four percent of 86 surveyed customer experience professionals said that lack of cooperation across}

organizations is a significant barrier to improving their organization’s customer experience. Source: Q4 2011 Global Customer Experience Peer Research Panel Online Survey.

7 _{Personalization technology has grown well beyond greeting people by first names into complex, real-time}

decision engines that make offers. But marketers still rely on stated demographic attributes to execute personalization because of the easy availability of this data in a customer profile. One retail bank executive told us that the customer’s post-login experience on the bank website consists of addressing the customer by name and including a message or offer based on the customer’s account profile — a fairly basic execution of post-login personalization. See the February 17, 2012, “Use Customer Analytics To Get Personal” report.

8 _{Source: Natalie Zmuda, “Home Depot Localizes With Spring-Loaded Pitch,” Advertising Age, April 4, 2011}

(http://adage.com/article/news/home-depot-localizes-marketing-suit-regional-weather/149699/).

9 _{Commonwealth Bank of Australia uses augmented reality to allow users to simply hold up a mobile device}

with the application open and point it at a home to get values that appear on top of the image. To see the prices relative to other homes in a neighborhood, a consumer simply moves his or her mobile phone so that the home is visible on the phone’s display. This action has fewer steps than typing in addresses or listing numbers and searching. See the December 22, 2011, “Case Study: Home Buying With Mobile Augmented Reality” report.

(12)

10 _{Forrester evaluated 10 WCM products across approximately 115 criteria and found that they have a lower}

level of maturity when viewed against these capabilities. SDL and Adobe lead due to their rich functionality, strategy, and enterprise track record. Strong Performers Sitecore, FatWire, OpenText, and Autonomy all provide strong options. Contender Ektron has solid functionality, but needs to work on its enterprise-level track record. Contender IBM lacks the broad set of capabilities others have in WCM, but has other CXM software assets and a strategic integration road map. Contenders Microsoft and Oracle have less of a CXM focus, but are legitimate alternatives for less interactive experiences. See the July 13, 2011, “The Forrester Wave™: Web Content Management For Online Customer Experience, Q3 2011” report.

In Forrester’s 75-criteria evaluation of global business-to-consumer (B2C) enterprise-class commerce suite vendors, we identified the 10 most significant software providers — Demandware, Digital River, hybris, IBM, Intershop, Magento, Micros-Retail, Oracle (ATG), RedPrairie, and SAP — in the category and researched, analyzed, and scored them. This report details our findings about how well each vendor fulfills our criteria and where they stand in relation to each other to help eBusiness and channel strategy professionals select the right partner for their B2C online and multichannel commerce solution. See the September 24, 2012 “The Forrester Wave: B2C Commerce Suites, Q3 2012” report.

11 _{It’s challenging to incorporate taxonomy projects within the whirlwind of tools and processes to create,}

manage, and access content in the enterprise. Organizations need a guide to building taxonomies that don’t sit on a shelf but work inside key information systems to make them hum. See the November 17, 2008, “How To Build A High-Octane Taxonomy For ECM And Enterprise Search Systems” report.

(13)

Forrester Focuses On

Application Development & Delivery Professionals

Responsible for leading development and delivery of applications that support your company’s business strategies, you also choose technology and architecture while managing people, skills, practices, and organization to maximize value. Forrester’s subject-matter expertise and deep understanding of your role will help you create forward-thinking strategies; weigh opportunity against risk; justify decisions; and optimize your individual, team, and corporate performance.

informs better decisions, and helps the world’s top companies turn the complexity of change into business advantage. Our research-based insight and objective advice enable IT professionals to lead more successfully within IT and extend their impact beyond the traditional IT organization. Tailored to your individual role, our resources allow you to focus on important business issues — margin, speed, growth — first, technology second.

foR moRe infoRmation

To find out how Forrester Research can help you be successful every day, please contact the office nearest you, or visit us at www.forrester.com. For a complete list of worldwide locations, visit www.forrester.com/about.

client suppoRt

For information on hard-copy or electronic reprints, please contact Client Support at +1 866.367.7378, +1 617.613.5730, or [email protected]. We offer quantity discounts and special pricing for academic and nonprofit institutions.