PDF File: 

CRM3_wp_chatham.pdf

Forrester Research

Firms expect online services to cut costs. But they won’t see savings unless Web-based solutions are aimed at receptive customer segments.

Firms plan to increase their spending on a host of new online service capabilities, ranging from natural language search to interactive knowledge bases. What's driving these investments? Interviewees expect to offload nearly one-fourth of their call volume to online self-service channels. In the wake of this e-service surge, firms struggle to manage customers that flow between on- and offline channels.

Firms Are Enamored with Online Service

Most large firms have deployed components of online service — both self-service and assisted-service activities like chat, email, and cobrowsing. To gauge the status of these offerings, we interviewed executives from 40 Fortune 1,000 firms with sizable call center operations.¹ Given their satisfaction with early efforts, respondents plan to augment basic capabilities like email and FAQs with more advanced features like interactive knowledge bases and natural language search.

"This has been successful — customers can view accounts, get information on policy values, cash values, and beneficiaries online." (Insurance company)

"Online order status was installed about a year and a half ago, and it significantly cut the number of calls in the call center." (Retailer)

Expectations Soar for Web Self-Service

Customer service executives expect a lot from Internet self-service — 58 percent of our interviewees describe this effort as critically important. What's driving this enthusiasm? Firms expect to offload 23 percent of their call volume to Internet-based self-service.

"Internet-based customer service technology will provide more self-help and faster problem resolutions. This will help the enterprise because each traditional contact that customers have with us is quite expensive." (Retailer)

"I want to get to the point where 80 percent of the time we can help customers through either Internet-based self-service or on phone-based self-service with the last 20 percent escalating to either chat or an 800 number." (Retailer)

Funding Increases but at a Slower Pace, Temporarily

Several respondents expect to hold back on cash outlays this year to digest significant investments in online customer service made last year. But overall, 61 percent of respondents expect to maintain or increase spending on e-service this year, and 50 percent anticipate budget increases in 2004.

"All our infrastructure was put in place last year, so this year we’re just finishing it up and leveraging our capital expenditure." (Telecom company)

"If we can identify good e-service technology solutions, we’ll spend whatever we need to. Our priority is customer-facing Web apps. They’re much more effective — and so much more cost-effective." (Transportation company)

When it Comes to E-Service, Firms Snub CRM Suites

Sixty-one percent of firms report that their e-service technology infrastructure ties together software from a variety of vendors. Respondents view large CRM suites as too pricey and inadequate compared with best-of-breed functionality.

"It would be ideal to have a voice recognition interface to natural language search linked to a knowledge base that can also be accessed through the Web. No vendor has all of the pieces." (Electronics company)

"We won't invest in a CRM suite vendor — that's counterculture for us. There's just no vendor that's best-of-breed in all applications." (Transportation company)

Organizational Change Remains a Big Hurdle

As respondents deploy e-service, they often find that securing internal buy-in to new online customer processes can be difficult.

"The biggest challenges to date have been internal. It's an extreme culture change for us to use the Internet as a customer service channel. So the development of processes to handle online feedback has been very slow." (Telecom company)

"With our online capabilities for checking in and printing boarding passes, it's been difficult to get airport agents’ buy-in and ease their security fears." (Airline)

Firms Struggle to Track Customers Across Channels

Respondents use a variety of metrics to measure e-service effectiveness — from usage volume to customer satisfaction surveys. But interviewees struggle to quantify the impact on customer interactions that cut across multiple channels.

"Customers who used to track their shipments by phone were checking the status approximately 1.1 times per shipment. When we introduced automated tracking on the site, that number went up to 8.6." (Transportation company)

"We can't prove that an electronic transaction replaces one over the phone. We’re worried that customers are using all of our channels to get the same service issue resolved." (Telecom company)

Don't Deploy Technology — Solve Problems

While companies hope that online service will reduce costs, they mistakenly scrutinize individual interactions. To succeed with e-service, firms must target specific customer problems, combining Web implementations with a sustained campaign to wean malleable customers off agent-based assistance.

Figure 1 — The Lifecycle of a Customer Issue

E-Service Optimism Needs a Reality Check

It's not surprising that companies pine for online customer service, especially when they expect self-service interactions to displace costly calls with agents. But even the e-service vanguards struggle to quantify the impact of this new channel. What's the problem? Firms track the Net as an isolated channel, while customers blend Web interactions with contacts in other channels (Figure 1). Although firms assume that phone-based representatives are the most expensive channel, our analysis of variable costs shows that this may not always be the case — especially when (Figure 2 and Figure 3):

• Self-service generates costly phone follow-ups. Despite offering a free initial contact, self-service turns out to be costly if customers unsuccessfully muddle around a Web site. Disillusioned patrons will either pick up a phone or give up. If the inbound call isn't handled well, the caller's pent-up frustration will increase the need for a supervisor's intervention. The result is that fewer issues get resolved and incoming phone calls get more costly. In our model, a poorly performing self-service effort can end up costing $6.55 per resolved issue — $0.38 more than interactions that began on the phone.
• Chat ties up reps — online and then on the phone. Depending on the rep's effectiveness, a chat-initiated issue can cost between $1.69 more and $1.87 less than a phone call. Although agents handle simultaneous chat conversations with customers, in most situations they can't deal with any more than three or four of these dialogues at a time. And lacking the verbal cues that help closing off phone calls, chat sessions often go on and on. Given the real-time nature of these interactions, customers may quickly give up and move to the phones.
• Email takes multiple exchanges to reach resolution. While email management tools help agent productivity, it remains difficult to resolve issues via email on the first interaction. Without being prompted in a dialogue, customers often don't provide all of the necessary information. The impact: Email conversations can span multiple interactions, often ending up with the customer making a phone call if the problem isn't resolved. That's why email costs can vary so wildly — from $4.78 to $9.53 for each resolved customer issue.

Cross-Channel Customers Need Integrated E-Service

Rather than just randomly deploying isolated e-service technology, companies must develop a broader service strategy that looks at customer interactions:

• By persona. Do customers want online service? Some do, and some don't. That's why smart firms must target offerings at specific personas — segments of customers with common goals, needs, and preferences (see the August 14, 2001 Forrester Brief "Scenario Design Depends On Personas")². When looking at online chat, for instance, firms may draw in some young male patrons, but they shouldn't count on attracting too many seniors.
• Across multiple channels. Customers freely bounce across channels, using a collection of Web, email, and chat interactions to solve their problems. So firms can't develop e-service offerings without looking at how their efforts will blend together with other on- and offline contact points. In our recent research, Forrester found that email and telephones remain dominant service channels — even for consumers trying to resolve issues with an online purchase (see the March 1, 2002 Forrester Brief "Chat Gains Ground As A Service Channel")³.
• Throughout an issue's lifecycle. Customers don't contact companies to build relationships — they’re looking to accomplish something much more tangible like finding an account balance or locating the replacement for a broken part (see "Build Customer Experiences, Not Relationships," March 1, 2002 Forrester Brief).⁴ That's why firms can't just examine individual customer touch points like a phone call or an email request. They must monitor interactions from the customer's point of view — from the inception to resolution of an issue.

Making E-Service Work

A clear plan will help companies avoid misguided e-service projects, but how can companies maximize success? Based on discussion with both users and e-service vendors, Forrester identified three areas of best practices:

1. Condition customers to use the "right" channels. While firms can't force customers online, they must coax malleable customers into low-cost channels.
2. Identify and preempt problems. Rather than just reacting to customer issues, companies should seek ways to avoid problems before they occur.
3. Deliver smart content. Customers won't use online service unless it addresses their ever-changing needs.

Condition Customers to Use the "Right" Channels

Two common fallacies drive many e-service failures. First, build-it-and-they-will-come efforts often end up with customers exercising their right not to come. Equally mistaken is an altruistic passion for providing equal access for all customers to every channel. While companies shouldn't arbitrarily shut off their 800 numbers, they should nudge customers toward low-cost interactions by:

• Marketing self-service. If there's a self-service capability that works, then don't be shy about encouraging customers to use it. How? With a combination of good site design, offline promotions, and usage incentives (Figure 4)
• Monitoring and adjusting the campaign. Is an e-service effort working or not? That's an unanswerable question unless a company sets specific milestones. Companies must treat new service channels like product rollouts — scrutinizing who's using them as well as how effectively they meet user needs. One company wasn't getting the traction it expected, so the firm tried little tweaks, such as changing the wording on the link to self-service from "support" to "help." The result: Online service usage nearly doubled.
• Blending escalation into the process. After spending 20 minutes unsuccessfully hunting for a replacement part at W.W. Grainger's Web site, a technician calls the distributor's standard 800 number. The service rep knows nothing about his online efforts but readily identifies the necessary part. What just happened? The customer has been conditioned not to start online. To avoid this Pavlovian effect, the move to agent-assisted service should appear as an extension to the online experience. Providing online self-helpers with a separate call queue or scripting an acknowledgement from the representative can go a long way.
• Turning representatives into online advocates. Who has the most understanding of customer issues? Representatives. Companies must get these frontline service providers involved in their efforts. Smart firms involve their agents in activities like identifying opportunities for self-service and informing callers of Web capabilities. Lands’ End could run a two-month contest to see which representative gets a higher online usage from their callers, using the lessons learned to institutionalize practices for targeting the right customers with the right messages.

Figure 2 Hypothetical E-Service Scenarios – By Customer Channel Used for Initial Contact

Identify and Preempt Problems

Leading companies don't just lower the costs of inbound requests, they look to avoid some customer-initiated issues completely by:

• Finding the root cause of "bad" calls. Customer contacts may be expensive, but they’re not necessarily avoidable — clients will always need access to help. But firms must identify customer contacts that flow from problems, such as broken parts or complaints. For example, rather than just automating the process for collecting complaints on delivery problems, the U.S. Postal Service should identify and resolve high-impact problems — like spending more of its e-service budget on analytics and problem resolution workflow.
• Proactively communicating with customers. While e-service efforts mostly deal with inbound requests, firms should also consider proactive communications. ZOOTS, the dry-cleaner firm with a strong home-delivery business, uses outgoing emails and prerecorded phone calls to remind patrons of scheduled pickups. Before turning off consumer electrical service, Puget Sound Energy uses PAR3 Communications to proactively call nonpayers. The system automatically identifies and calls target customers without even involving an agent. The result is a 19 percent increase in collections.
• Monitoring the customer environment. Hewlett-Packard can remotely diagnose deployed servers, identifying root causes of problems and notifying customers about potential issues like security vulnerabilities and disk space shortages. This type of proactive service will increase as more manufacturers start selling sensor-embedded products (See the Forrester Report "The X Internet Invigorates B2B Apps," October 2001.)⁵ Look for Maytag to alert customers when a thermostat or washer arm is likely to break.

Figure 3 — Costs for 100 Customer Service Issues

Deliver Smart Content

To keep customers hooked on e-service, firms must ensure that online users consistently find answers and resolve problems. How? By institutionalizing processes that make sure that content remains:

• Accessible. Customers don't distinguish between answers that aren't on the site and those that they can't find — it's poor service either way. Burying self-help on the Web or throwing a lot of content at users will neutralize even the best capabilities. Firms must tailor offerings to specific types of problems. To ensure that users get easy access to whatever help they need, companies should perform scenario reviews, evaluating usability based on specific user goals (See the May 2002 Forrester Report "Best Practices For Web Site Reviews.")⁶
• Consistent. Customers won't trust Target's self-serve Web site if it shows three retail sites with pharmacy operations in the San Francisco area while call center representatives can list off five. To avoid this mismatch, firms should consolidate content within a single knowledge base for both internal and external usage from a vendor like KANA, Primus, or ServiceWare — or synchronize disparate sources if the organization won't tolerate centralized systems (see Forrester Report "Web Services Erode CRM Barriers," May 2002).⁷
• Fresh. For firms with dynamic product lines, such as computer retailer Micro Warehouse and software tools vendor Rational, information can rapidly grow stale. That's why these firms need to continually update its collection of issues and resolutions. How? By institutionalizing a knowledge base maintenance effort within the normal workflow of the call center. A good first start: Define weekly targets for representatives in documenting, updating, and refining solutions. Firms should also plan on staffing editors for tweaking and approving content for publication.

Figure 4 — Hypothetical E-Service Scenarios by Customer Channel Used for Initial Contact

Companies planning to improve their online customer service should:

Create Customer Service Personas

To target the most effective prospects for self-service, firms must segment their customers according to their attitudes toward, and use of, technology. Where can you start? By using Forrester's Technographics® data (see the November 2001 Forrester Report "Why Technographics Works").⁸ With this segmentation in place, firms could develop detailed descriptions of a few typical customers in each key segment — a mechanism that Forrester calls personas. This archetype will help organizations visualize how to best service the segment.

Develop Pricing for Self-Servers

As companies master e-service, they’ll find a new type of customer: the self-server. These predominantly online stalwarts will cost less to serve than their phone-centered peers. To acknowledge these economics, companies must plan for at least three tiers of service offerings: 1) online only 2) limited phone access and 3) unlimited phone access. With these levels in mind, firms should develop mechanisms like tiered pricing or rebates to reward and encourage self-servers.

Initiate Smart Escalations

Since companies can't expect customers to use the right channel for their issues, they must prepare to switch interactions across channels. If a Fleet Bank representative gets an email from an important customer about an emergency wire transfer but the message lacks some of the required information, an alert should go off. Rather than starting a potentially long string of emails, the representative should call the customer — or email instructions on getting priority access in the phone queue.

Get CTI Now

As companies try to deliver cross-channel service, they’ll need to provide representatives with rich context about inbound callers — recognizing, for instance, if they’ve recently been online. But this type of cross-channel information requires computer telephony integration (CTI): a technology that is only deployed at 53 percent of our interviewees’ firms. If the other 47 percent hope to capitalize on self-service without degrading customer satisfaction, then they must plan to upgrade their handling of inbound calls ASAP.

What It Means

Call Centers Will Replace Hourly Wages with Salaries

Self-service improvements over the next few years will allow companies to offload a significant volume of simple questions coming into their call centers. The result: a higher percentage of complex inbound calls. To deal with this increasing sophistication, contact centers will revamp their staffing models, using highly skilled agents instead of low-price tier-one representatives as the initial contact point for inbound calls.

Firms Will Shut Off Email and Chat

What will happen to companies that can't master e-service? They’ll see their service costs increase while their competitors’ decrease. Rather than continuing to sink money into agent-assisted efforts, some companies will eliminate email and chat as options, returning to their heritage: a dependable phone call.

E-CRM Vendors Will Disappear

As companies increasingly try to serve customer issues — not just isolated touch points in different channels — integrated platforms like Siebel and mySAP CRM will look more appealing. The result: stagnation for mid-size Internet-centric CRM players such as KANA and eGain. Look for these online point plays to sell out to bigger vendors like SAP and PeopleSoft over the next 12 months.

Knowledge Base Vendors Will Become CRM Trophies

Quick access to answers represents the cornerstone of most e-service efforts. As companies focus more on developing and maintaining knowledge bases, knowledge solutions will become key elements of a CRM selection. Look for Siebel to buy Primus — and fold it into the software giant's offering.

Analytics Niches Will Get Rolled into Suites

Companies will increasingly want to track user experience across every aspect of their business — across channels, life cycles, and by different customer personas. But to handle that holistic analysis, firms must piece together offerings from players across niche areas like Web performance (e.g., TeaLeaf Technology), interaction efforts (e.g., NICE Systems), telephony queues (e.g., Empirix), and customer percent satisfaction (Satmetrix Systems). In this gap, look for systems integrators like iXL/Scient to expand their customer experience practices and for vendors like SAS to make acquisitions that build out full analytical suites. 

Endnotes

¹ All our interviewees had a minimum of 200 call center representatives, and the median size was 1,200 representatives.

² Personas — detailed descriptions of user archetypes that represent market segments — are critical to the process of Scenario Design. Companies that create design personas have the necessary foundation for crafting a great user experience.

³ Online buyers are more likely to use email and the phone to resolve their customer service issues — 89 percent and 57 percent, respectively — than other service channels. But beyond these dominant channels, alternative customer service solutions have emerged in the past two years.

⁴ Since customers will rebuff corporate relationship advances, firms need a new CRM focus. Forrester recommends that companies build and sustain their customer base by obsessing about customers’ experiences at every touch point.

⁵ Forrester foresees the emergence of the extended Internet that will exploit ever-cheaper sensors as well as smart tagging and tracing technologies to enable manufacturers to track every product they make from inception to phaseout.

⁶ Forrester believes that expert reviews are a powerful way to evaluate both software and Web sites. The key: Adopt best practices for "scenario reviews," which we define as checking an application for compliance with research-based rules while attempting to accomplish a specific user goal.

⁷ Firms can lower the overhead of sharing and synchronizing data like FAQs or order status if they start by presenting interfaces that other corporate business units can tap via Web services.

⁸ Forrester's Consumer Technographics® research defines segments based on three dimensions. The first is technology optimism: people who are convinced that technology will enhance their lifestyle are drawn to it, even if they are concerned about the effects of technology on society as a whole. The second is household income: Once we know how entranced by technology someone is, we look at whether she can afford to give in to her technology urges. The final dimension is primary motivation. This attitudinal dimension tells us why a consumer will choose one technology over another and suggests how he will use it.