Guidelines to cooperation

Editor’s note: Hernando Gonzalez is senior manager, customer research, Sun Microsystems, Louisville, Colo. Joan Fredericks is senior vice president, director of advanced methodology, Harris Interactive Loyalty, Princeton, N.J.

About three years ago, Harris Interactive Loyalty, a division of Harris Interactive, Rochester, N.Y., developed a statistical model of customer commitment based on business-to-consumer (B2C) interactions. This model demonstrated that both rational and emotional connections on the part of customers predicted their behavior. Quirk’s featured this customer relationship model in the article “Moving Beyond Satisfaction” in its April 2005 issue.

At this time, the customer research team at StorageTek, a data management and storage technology company, had been working with Harris Interactive Loyalty for several years and immediately saw the possibility of applying the commitment model to its business-to-business (B2B) environment. They collaborated in determining that the model could be applied to StorageTek’s global customer satisfaction and loyalty surveys. Two subsequent survey waves showed that, even in B2B relationships, emotional factors contributed almost as importantly toward likelihood to recommend as did rational factors.

StorageTek was acquired by Sun Microsystems as a business unit in 2005, and is now a major Sun brand. Following are 10 suggestions for a successful customer satisfaction and loyalty research program, based on collective research expertise, best practices and the B2B work conducted by StorageTek and Harris Interactive.

1. Prepare the organization.

2. Pay attention to the emotional side of customer relationships.

3. Address competitors’ strengths and vulnerabilities.

4. Segment customers based on their needs and value to the company.

5. Assess the effect of cultural differences on results.

6. Address channels as part of the customer research process.

7. Communicate and share lessons learned throughout the process.

8. Turn information into decision tools.

9. Create financial linkages.

10. Respond to internal and external needs for change over time.

1. Prepare the organization.

In order to succeed, a customer satisfaction and loyalty research program requires support from  senior and middle management from the very beginning. Buy-in from marketing and sales teams is often necessary, as they are implementers of survey recommendations and providers of up-to-date account contact information. A large part of this support has to be earned by the customer research team. One way to accomplish this includes using rigorous scientific methods and providing research findings that are actionable at both strategic and tactical levels. After all, the purpose of doing customer research is to provide practical tools to marketing and sales teams to enable them to retain their most profitable customers and continue to win market share among carefully targeted segments.

Another way to win commitment at the field level is to tie the customer satisfaction and loyalty results to a percentage of the annual bonus. Senior management’s support of this initiative signals to the field that the company is serious about continuously improving its satisfaction and loyalty scores by leveraging sound, performance-focused research.

2. Pay attention to the emotional side of customer relationships.

Typical customer satisfaction and loyalty research programs do a competent job of addressing functional performance. Surveys generally tend to focus on measuring product quality, service and price perceptions. What is often neglected, however, is the emotional side of customer relationships; yet the most committed relationships are based on creating both rational and emotional bonds with customers. The result? When customer relationships are strong on both of these dimensions, customers behave in more profitable ways: giving greater share-of-wallet, purchasing more frequently, being willing to try additional products and even being less price sensitive.

Emotional bonds are developed by creating trust. When customers trust a company, they believe it is acting as their partner and is looking out for their best interests. While trust may sound like a warm-and-fuzzy concept, it is built through concrete actions that encompass the way a company manages day-to-day touchpoints with its customers and policies, and the manner in which the company lives up to its corporate reputation and brand.

Here are some examples of specific drivers of trust and emotional connection, based on numerous statistical models built for B2B environments:

  • account representatives respond quickly to requests (touchpoint);
  • terms of warranty (policy);
  • behaves ethically (reputation).

While trust is critical to building an emotional connection with customers, the rational side of customer relationships is driven by satisfaction with functional performance. Satisfaction is achieved by consistently meeting product and service quality expectations and delivering excellent value for the cost. Examples of typical B2B satisfaction drivers include:

  • reliable products (product quality);
  • on-time delivery (service quality);
  • competitive pricing (cost).

Path models are useful in understanding complex relationships. The commitment model (Figure 1) shows how trust and satisfaction together created emotional and rational bonds with data storage managers. These bonds in turn explained customer commitment and behavioral intent or loyalty. The attributes in the rectangles are examples of the questions used to measure each construct in the model.

The emotional side of customer relationships in this environment proved to have a strong and significant impact on customer commitment and loyalty. Although the rational side had a somewhat greater weight, the ability to explain customer commitment and loyalty would have been much weaker if the model failed to identify and measure the emotional connection.

As expected, trust supported an emotional connection and satisfaction a rational one. In addition, there were some subtle indirect influences at work. Typically, satisfaction is created through consistently providing quality and value. In this model, satisfaction contributed to building trust. Trust in turn not only explained the emotional side of the customer relationship but also influenced the rational side. This meant that the more StorageTek acted as a partner to its customers (along with other trust-building behaviors), the more customers believed in the quality and value of its products.

The complete model contained detailed diagnostics, showing how improvements in specific products, services, pricing, reputation and other attributes affected trust, satisfaction and the emotional and rational connections with customers.

3. Address competitors’ strengths and vulnerabilities.

While customer loyalty research professionals might take this for granted, a surprising number of B2B companies focus their research solely on their own performance. They may go so far as to measure their customers’ ratings of competitors without obtaining a true benchmark based on competitors’ customers.

StorageTek developed a comprehensive research process, addressing its own customers as well as those of its competitors. This provided the business unit with a clear understanding of its competitive position, and the ability to leverage unique strengths to increase share-of-wallet and attract new business among targeted segments.

Ideally, a complete picture of customer loyalty will cover all of the components shown in Figure 2. An understanding of a company’s performance, its competitors’ performance and the impact on loyalty of each attribute will chart the course to identifying and achieving clear improvement priorities.

4. Segment customers based on their needs and value to the company.

Most B2B companies segment their market in some way. The segmentation may be based on firmographics such as size or revenue, verticals or some other characteristic. The shortcoming of this view is that the segmentations may be suboptimal. When companies use segmentation in building a model of customer loyalty, they need to ask themselves the following:

  • Is this the best way to segment the market?
  • Is the segmentation scheme valid?
  • Are the most profitable segments being targeted? Are we using the most effective strategies?

StorageTek recognized the importance of segmenting its market in line with these considerations. With Harris Interactive’s help, it conducted a segmentation study to assign customers and prospects to groups that shared commitment drivers (needs) and targetable characteristics. This enabled it to assess the value of each segment to its business and develop compelling value propositions required to attract and grow customers within the highest potential segments.

Figure 3 is a fictional example of the type of segments that can be uncovered through customer loyalty research.

5. Assess the effect of cultural differences on results.

When confronted with results of customer loyalty research, many have been heard to say, “My customers are tougher graders.” Customer research lore is full of anecdotal evidence of cultural or country differences in results. But how can one know whether these differences represent tougher-to-please customers or just differences in the ways customers use rating scales? And, even more importantly, how can one level the playing field when using customer loyalty research results to compensate managers who are responsible for different geographic, ethnic or cultural groups?

This becomes especially complicated in today’s flat world, where a given customer may have been born in India, educated in the U.S. and is currently working in Singapore. Just applying a country-level “correction” fails to capture the individual differences behind the research results. Different individuals use scales in different ways. For instance, on a satisfaction scale of zero to 10, one person gives scores centered around the midpoint, so a score of eight is exceptionally positive coming from  this person. Another gives scores that are relative to the positive endpoint. In this case, 10 is this person’s most common score, so a score of eight is a sign of disappointing performance.

Figure 4 shows how a score of eight from  different types of individuals might be interpreted based on their usage of a zero-to-10 scale.

When these individual differences are rolled up at a country or regional level, they can have a substantial impact on results. Figure 5 shows country-level differences from  the European average, based on ratings of hundreds of products using a zero-to-10 rating scale.

When communicating their customer loyalty research results, StorageTek faced similar questions about culturally-based differences from  its regional and country managers. In the particular approach that Harris Interactive used for StorageTek, respondents were asked a brief series of rating questions across a wide variety of products and services. Their evaluations of these carefully selected brands uncovered inherent patterns of scale usage at the individual respondent level. These patterns were then used to apply an individual correction to remove noise, in the form of response bias, from the data.

Depending on the range of responses and the available respondent profiling information, a variety of approaches can be taken, some without the need for additional cross-category ratings. The important point is that, while some types of respondents are more homogeneous in their scale usage than others, assumptions about scale usage need to be tested for the specific customer population at hand. Results can then be adjusted as needed, so that results from  country to country or among other subgroups of interest are comparable.

6. Address channels as part of the customer research process.

B2B markets often rely on fairly complex channels: value-added distributors (VADs) and value-added resellers (VARs), as well as traditional distributor and retailer relationships. Depending on the industry, an indirect customer may or may not even be aware of which company’s product or service they are purchasing. B2B companies are dependent on their relationships with these channels, as well as the relationships the channels have with their customers. Best practice customer loyalty research calls for addressing these channels both as partners and as customers in their own right. Figure 6 illustrates channel/partner “triangulation” in B2B relationships with customers.

StorageTek and Harris Interactive understood the critical role of channel partners and designed the customer loyalty research process to address channel/partner triangulation. This led to a full understanding of how customers who purchased indirectly were affected by channel relationships as well as StorageTek’s own performance.

7. Communicate and share lessons learned throughout the process.

The most tragic event that can befall a customer satisfaction and loyalty research program is to end up with reports that no one reads or cares about. Consultation with business unit managers and teams in different geographies begins with the preparation of the survey questionnaire. In the end, every research question has to be tied to a business objective.

The consultative process continues over time, with every survey iteration. As each research wave comes to an end, the customer research team needs to interpret, analyze and share findings and recommendations at different levels. Sharing does not have to end with meetings and Webinars. The customer research team can build an internal Web site (portal), with different levels of security clearances, to enable marketing and sales teams, as well as members of senior management, to access previous reports, and even data, for additional analyses.

8. Turn information into decision tools.

Customer research teams need to convert research data into meaningful findings and recommendations. If the internal stakeholders are adequately consulted on the front end of the research, there should not be any problem in understanding and using the research when the analysis and recommendations are submitted. Customer research teams that fail in this task very often do not identify the most important business objectives and priorities of business units and regional or country managers. As a result, these managers and their teams may not find anything useful or relevant in the reports.

Simplifying complex statistical reports is not an easy task - it requires a different way of communicating at various levels of the company. Reporting at the senior management level very often requires global overviews and empirically-based generalizations, while communicating at the sales team level means providing them with factual details relevant to their specific conditions, target customers and geographical areas.

Harris Interactive turns its statistical models for clients into desktop simulators so that management can play what-if games to estimate the impact of specific improvements on customer loyalty. These simulators can incorporate the effect of planned organizational initiatives to improve customer loyalty.

9. Create financial linkages.

Customer satisfaction and loyalty does not mean keeping every customer in every case. For example, in those industries where 80 percent of a company’s revenue comes from  only 20 percent of its customers, it makes sense to identify which customers belong to the 20 percent that matters. Without question, those are the customers that a company must delight and whose expectations they must exceed.

On the other hand, how much of the remaining 80 percent of its customers should a company keep loyal? How profitable are those customers? This may require further segmentation by level of profitability, because some customers may simply not be worth the effort. If the company loses money or just breaks even on every transaction, efforts should be made to convert these customers into more profitable customers. If that is not possible, then the company has to consider whether keeping those customers is worthwhile. This is the true test of any customer loyalty research program. Trying to keep every customer may, in the end, be a great disservice to the company.

10. Respond to internal and external needs for change over time.

The search for relevance never ends. Market conditions change. Competitors and substitute products or services alter the landscape. Marketing would be too easy if a company were always first with disruptive innovations or leading-edge products. But that scenario rarely happens. Customer research teams need to be in tune with what is happening in their industry, in the key verticals that their company serves, and with their customers’ unmet needs and wants.

The customer satisfaction and loyalty survey questionnaire has to change over time to reflect these conditions. The key changes may be small from  one year to the next, since comparability between time points is an important consideration. Over a longer period of time, however, those cumulative changes would make the survey questionnaire a more sensitive instrument. Some questions may change in meaning, or may no longer be relevant, given certain developments in the market.

Again, suggestions from  customers, marketing and sales teams, and managers in close touch with industry developments and conditions are invaluable resources. Listening to customers and learning from what they would like to do, but are unable to do, with current products or services will open up new opportunities.

Critical component

These are just 10 of many best practices for developing and implementing a successful B2B customer satisfaction and loyalty research program, based on past experiences of StorageTek and Harris Interactive. Building customer loyalty and retaining profitable customers is a critical component to any company’s prosperity. There are many factors to consider in the design of a comprehensive customer satisfaction and loyalty research program, particularly the emotional side of customer relationships. By strengthening both rational and emotional bonds in its business relationships, StorageTek, as part of Sun Microsystems, stores more digital archival data than anyone else (Freeman Reports: The Growing Importance of Archive, May 2006 http://www.freemanreports.com/pdf/STKArchive.pdf ).