You have to measure it to avoid it

Editor’s note: Al Fitzgerald is president and founder of Answers Research, Solana Beach, Calif.

The typical goal of satisfaction and loyalty research has been to identify and track the percentage of a brand’s loyal customers with the hope of increasing the percentage of secure customers and reducing the percentage of at-risk customers. This is usually done by asking questions in three key areas: 1) overall satisfaction, 2) likelihood to repurchase and 3) likelihood to recommend. Secure customers are defined as those that report high scores on likelihood to recommend, likelihood to repurchase, and overall satisfaction. Neutral customers give middle ratings on at least some of the three questions. At-risk customers give a low rating on any of the questions. The strategy behind this has been to identify and track the percentage of customers who fall into each loyalty category with the hope of increasing the percentage of secure customers and reducing the percentage of at-risk customers.

It is a major assumption that a secure customer will continue to buy, or at least have a strong affinity to, the brand even if a competitor lowers price or improves features.

The typical definition of a secure customer is flawed

High ratings on likelihood to recommend, repurchase, and satisfaction are correlated with loyalty scores - but they do not predict loyal behavior. Loyalty is the propensity, in the short term, to continue to purchase a brand even if competitors’ products are less expensive, have better availability or have better features. Essentially, true loyalty means that a customer will make an uneconomic decision - they will stand by a brand and continue to purchase it even if it is more expensive or has been passed by technologically or feature-wise.

While many loyalty studies are able to show high correlation coefficients (R-square) with loyalty scores, these do not prove causality. The high R-square stats typically indicate a high correlation - but what is correlated is not necessarily causal. Therefore, while typical loyalty studies are good at predicting loyalty scores they are much poorer at predicting loyal behavior. That is, when a brand’s price or feature set becomes temporarily non-competitive, the brand’s secure customers buy from the competition. Clearly these secure customers are not very loyal after all.

Here is a fictional example: Let’s say there is a Mobil gas station near a customer’s home. It is very clean and convenient. A customer always buys gas there. He gives this station the highest marks on likelihood to recommend, likelihood to repurchase, and overall satisfaction, typical questions used to assess loyalty. However, if a Texaco station opens across the street and its price is one cent cheaper, this particular customer switches to Texaco!

That is why we need to know more than the standard three questions when assessing satisfaction and loyalty.

What defines true loyalty?

To understand true loyalty, we must understand what motivates a person to buy or use a brand or service. All customers can be segmented into one of the following three groups: price-driven, feature seekers, or brand loyalists (Figure 2).

Each group has its own motivation for purchase:

• Price-driven customers. Their purchases are based on price primarily. They seek a good deal. They are more influenced by price differences than by feature or brand differences in a product.

• Feature seekers. They are driven to buy the “latest and greatest” and differentiate between products based on concrete, tangible features. They are often willing to pay a premium to get a fully-featured product and may be unwilling to compromise for a brand that does not have the specific features they seek.

• Brand loyalists (intangible feature seekers). Often standardize on one brand. These buyers perceive that it is the brand that is important and will delay a purchase if their preferred brand is temporarily unavailable. They are often less price sensitive and will purchase name brands, even at high price premiums. Brand loyalists will classically buy their preferred brand even if the brand is lagging behind competitors on features/service specifications. These buyers believe that their brand will eventually catch up and are not highly influenced by the “latest and greatest” whiz-bang feature.

Only brand loyalists can be loyal to a brand

In our gas station example, we saw that a secure customer who rated the Mobil station high would immediately switch to Texaco if Texaco’s price was one cent lower. Why? The customer is price-driven. This customer likes the Mobil station well enough. But when a competitor with acceptable features and lower prices arrives, the price-driven customer switched brands. Price driven customers are not loyal to brands! Likewise, feature seekers are not loyal to brands! Since many early adopters seek out the latest and greatest gadgets, they seek out the best features. If a new product enters the market with an improved feature set, the feature seekers will immediately switch to the brand with the best features. Feature seekers often agree with statements such as “all companies are alike.”

Only the brand loyalists are loyal to brands. Why? Brand loyalists actually do seek features, but they seek intangible features. The intangible features of a brand are extremely important to these customers. Often times these intangible features include reliability, compatibility, trust, service and support, perceptions of the firm’s financial stability, or that the firm is “an industry leader.” Many of these intangible features (which brand connotes) overcome the “FUD factor” (fear, uncertainty and doubt). Essentially, “brand” is a surrogate for a basket of intangible features which are difficult to independently assess. Therefore, these customers rely on the brand’s reputation when making purchases.

Know them, retain them

Many firms do not know their customers’ loyalty traits and make mistakes when seeking to enhance loyalty. For example, the Mobil station may try to retain its secure customers by planting nicer flowers, having cleaner restrooms or providing more additives in its high-end fuels. However, none of these actions will increase loyalty for our example customer. Instead, Mobil, if it knew that this customer was price-driven, would use these funds more wisely by lowering the price of its fuel. This would make our example customer less likely to switch brands. Some customers will put up with bad service, poor quality or inconvenient hours and continue to purchase long-term. They continue to purchase because these characteristics are less important to them than price or features.

Actions that will retain price-driven customers will be different than actions needed to retain brand loyalists or feature seekers. If a firm were to discover that 10 percent of its customers were unhappy (i.e., low ratings on recommend, repurchase and/or satisfaction) what strategy should it employ? The strategy depends entirely on who is unhappy:

  • If the unhappy customers are primarily the price-driven customers, there is much less to worry about because these customers are not loyal anyway and will likely continue to buy as long as the brand maintains a competitive price.
  • If the unhappy customers are feature seekers, there is less to worry about unless somehow the product is not delivering a promised feature/service. Remember that feature seekers will switch brands on their next purchase if another brand offers a more attractive feature set. Strategically, a firm should allocate its budget to stay on the cutting edge of features in order to keep feature seekers.
  • If the unhappy customers are disproportionately brand loyalists, then it is a more serious problem. Brand loyalists desire the intangible features that brand connotes. If a customer is dissatisfied, it is likely that this results from the brand’s inability to deliver on a key intangible feature - such as reliability, compatibility, or service/support. Strategically, it is critical to identify the offending area and make quick corrections to retain these loyal customers.

Dissatisfaction studies, not satisfaction studies

We have seen above that it is very important to understand which type of customer is dissatisfied. But many satisfaction studies survey too few dissatisfied customers to help us make the right decisions!

For example: Let’s say a typical satisfaction study lasts 20 minutes with 400 customers surveyed. In a typical study, roughly 10 percent of customers are dissatisfied. The percentage is rarely higher than 10 percent because dissatisfied customers quickly become non-customers and are excluded from future surveys. In this example, we would only speak with 40 unhappy customers and this may not identify all important areas affecting satisfaction.

Instead, we should focus on surveying dissatisfied customers. All studies have a limited budget. Instead of surveying 360 satisfied and only 40 dissatisfied customers, it would be a better use of interviewing dollars to focus on those who are NOT satisfied. We suggest using overall satisfaction as a screening question. We track the percentage who are satisfied, overall, to see if this percentage changes over time. However, we only ask additional questions to those who are dissatisfied. Typically, for the same budget, we can find 250-300 dissatisfied customers rather than 360 satisfied and only 40 dissatisfied. This approach dramatically increases the number of dissatisfied customers who are surveyed. Therefore we gain substantially more detail on how to avoid future dissatisfaction problems. We also are able to understand who is dissatisfied - the brand loyalists, the price-driven or the feature seekers. After all, unless we know who is dissatisfied, we are not able to allocate our limited resources optimally.