Customer Satisfaction Metrics: From Touchpoints to Journeys

A March 2016 article in McKinsey Insights and Publications addresses an interesting problem about customer satisfaction. I explore a gap in the article's coverage and highlight some interesting results that can be used to better engineer the customer interaction experience.

Typically a customer's experience of a company is built up from interacting with a number of different customer service teams over a period of time. For example, to set up my cable box I first activate the device via the company's website, call the technical support line when I have a problem, call sales to upgrade my content package, and chat online with billing to clarify my first month's bill. These individual interactions or touchpoints together constitute my customer experience journey.

But here's the problem.
"Consider the dilemma that executives faced at one media company. Customers were leaving at an alarming rate, few new ones were available for acquiring in its market, and even the company's best customers were getting more expensive to retain. 
...While the company's overall customer-satisfaction metrics were strong, focus groups revealed that a large number of customers left because of poor service and shoddy treatment over time. "How can this be?" one executive wondered. "We've measured customer satisfaction for years, and our call centers, field services, and website experience each score consistently over 90 percent. Our service is great!" 
As company leaders probed further, however, they discovered a more complex problem. Most customers weren't fed up with any one phone call, field visit, or other individual service interaction -- in fact, most customers didn't much care about those singular touchpoint events. What was driving them out the door was something the company wasn't examining or managing -- the customers' cumulative experience across multiple touchpoints, multiple channels, and over time. 
Take new-customer onboarding, for example, a journey that spanned about three months and involved an average of nine phone calls, a home visit from a technician, and numerous web and mail interactions. At each touchpoint, the interaction had at least a 90 percent chance of going well. But average customer satisfaction fell almost 40 percent over the course of the entire journey. The touchpoints weren't broken -- but the onboarding process as a whole was."
As customers move from one touchpoint to another, they may be satisfied at each touchpoint but still be unsatisfied by the journey as a whole. As the media company executive asks, "How can this be?"

How Silos Can Be Great While Journeys, Well, ... Not So Great


Figure 1: A simple picture of 6 customers, each progressing through 3 touchpoints

Figure 1 shows customers moving left to right through touchpoints on the way to completing their journeys. At each touchpoint in his or her journey, a customer is either satisfied or not satisfied by the service received at the touchpoint. For each touchpoint we can define the touchpoint satisfaction rate by taking all satisfied customers who have passed through the touchpoint and dividing by the total number of customers who have passed through that touchpoint.

At the end of the journey, a simple way to measure a customer's satisfaction with his or her journey as a whole is to take the number of touchpoints at which the customer is satisfied and divide it by the total number of touchpoints the customer navigated to complete the journey. Let's call this measure the baseline journey satisfaction rate. Each customer will have a baseline journey satisfaction rate which depends on whether or not they're satisfied at each of the touchpoints.

Of course there's another way to measure satisfaction at the journey level -- just ask the customer to rate the journey as a whole. Let's call this the refined journey satisfaction rate. But, as the authors of the article observe, most organizations never measure the refined journey satisfaction rate. Customer service is structured as a series of interactions with the customer, each interaction conducted with a siloed customer service team. At the end of the journey, no one is responsible for circling back with the customer and measuring the refined journey satisfaction rate.

From the customer service model shown in figure 1, we have three distinct types of customer satisfaction metrics: the touchpoint satisfaction rate, the baseline journey satisfaction rate, and the refined journey satisfaction rate.

As the authors explain, even when the touchpoint satisfaction rate for every touchpoint is high (say 90%), a given customer's refined journey satisfaction rate can be much much lower -- as low as 60%! (I'm not sure how the authors arrived at this number -- it could be estimated from the churn rate and a few other assumptions, but they don't say.)

There are some key reasons for this divergence between the touchpoint satisfaction rate and the refined journey satisfaction rate. Customers might be very satisfied at each touchpoint but unsatisfied by the journey if the touchpoints taken together were too few to solve their problem; they got a partial solution but not a complete one. Another scenario is when there are too many touchpoints; customers will be unsatisfied by the journey as a whole because they were put through the ringer. Yet another reason might be the lack of context or continuity in their interactions with the individual touchpoints; customer service silos might not pass on the relevant information to other silos downstream, making the customer experience much more trying and impersonal than it should be.

These are the troubles companies have when they set up customer service through a set of siloed touchpoints and the article's authors go on to provide some good advice on how companies, while sticking to their customer service silos can still manage to improve refined journey satisfaction rates.

There is another divergence that's briefly tackled -- McKinsey director Alex Singla explains it in a video that is part of the article. This is the divergence between the touchpoint satisfaction rate and the baseline journey satisfaction rate. Due to mathematical reasons alone, baseline journey satisfaction rates can be much lower (and, surprisingly, also higher) than touchpoint satisfaction rates. The particulars depend on the number of touchpoints and the pattern is interesting.

In what follows, I'll show you the results of studying this divergence between touchpoint satisfaction rates and baseline journey satisfaction rates in some detail. It's worth understanding because even if you solved all the problems with the overall customer journey, this problem would still remain. It's therefore useful to understand it and use this understanding to inform the way in which the siloed customer service teams should be organized -- in particular, how many silos to have, the drops in baseline journey satisfaction rates to watch out for, and what touchpoint satisfaction rates to aim for.

A Simple Case: Just Two Touchpoints


Say we have 1000 customers and they each pass through just 2 touchpoints. The touchpoint satisfaction rate for both touchpoints is 90%. What do you expect the baseline journey satisfaction rates to be? You might be surprised!

Figure 2: Distribution of customer journeys across baseline journey satisfaction rates for 2 touchpoints

The results in figure 2 are based on sending 1000 customers through 2 touchpoints with each touchpoint having a 90% touchpoint satisfaction rate. For good measure, I simulated this 200 times. For each simulation, I calculated the percentage of customer journeys that had a baseline journey satisfaction rate of 96% to 100%, 91% to 95%, and so on all the way to 0% - 5% (figure 2 begins displaying these rates starting at 41% to 45%). I then took the average percentage of customer journeys (shown on the vertical axis in figure 2) at each baseline journey satisfaction rate (the horizontal axis in figure 2).

Are you surprised that even when the touchpoint satisfaction rates are each 90%, on average, 18% of the customer journeys have a baseline journey satisfaction rate between 46% and 50%? That's almost 1 in 5 customers with appallingly low baseline journey satisfaction rates. And even though no touchpoint satisfaction rate was above 90%, 81% of the customer journeys (on average) have baseline journey satisfaction rates of 96% to 100%. So bad news and great news.

It's easy to see why this happens. If a customer is satisfied at a touchpoint, mark that as a 1 and if he or she is dissatisfied at a touchpoint, mark that as a 0. With 2 touchpoints, there are only 4 possible journeys any customer can take -- 00, 01, 10, 11. With a 90% chance of getting a 1, the 11 result dominates -- it occurs 81% of the time. But even with a 90% chance of getting a 1, there are enough 01 and 10 journeys such that the baseline journey satisfaction rate for 18% of the journeys is just 50%.

And this happens no matter how well a silo manages its customers or the various improvements a silo has made in treating its customers by providing context, the right feedback to other silos, and so on. At the end of the day, the simple mathematical fact is that both silos can have a 90% touchpoint satisfaction rate, but end up treating 18% of customer journeys badly!

What happens when we have 3, 5, 10 or more touchpoints?

Baseline Journey Satisfaction Rates for Multiple Touchpoints


Figures 3, 4, 5, and 6 show what happens when we have 3, 4, 5, and 10 touchpoints respectively.

Figure 3: Distribution of customer journeys across baseline journey satisfaction rates for 3 touchpoints

Figure 4: Distribution of customer journeys across baseline journey satisfaction rates for 4 touchpoints

Figure 5: Distribution of customer journeys across baseline journey satisfaction rates for 5 touchpoints

Figure 6: Distribution of customer journeys across baseline journey satisfaction rates for 10 touchpoints

A number of things jump out in this series of figures.

  • Even though the touchpoint satisfaction rate is always 90%, only the case in which there are 10 touchpoints registers baseline journey satisfaction rates that are in this range (86%-90%).
  • Even though the touchpoint satisfaction rate is always 90%, all the above scenarios have a significant percentage of customer journeys with baseline customer journey rates above 90%. That's the good news. The bad news is this percentage of customer journeys drops steadily from an average of 73% in the case of 3 touchpoints all the way to 35% in the case of 10 touchpoints.
  • Having more touchpoints is not necessarily worse. What companies should be aware of is that for mathematical reasons alone, there will be customers who have baseline journey satisfaction rates that are much lower than any of the touchpoint satisfaction rates. For example, in the case of 4 touchpoints, it might make sense to identify the 29% of customer journeys that have baseline journey satisfaction rates between 71% and 75% and intervene to ensure customer retention. And similarly for the 10-touchpoint case, perhaps it will make sense to find and rectify the experience for those 6% of customers who have baseline journey satisfaction rates between 66% to 70%. Awareness is power.

What happens when there are even more touchpoints? The summary results are shown in the table in figure 7.

Figure 7: Mean % of customer journeys as a function of the number of touchpoints (rows 2 to 50) and the baseline journey satisfaction rate (columns 46%-50% to 96% to 100%)
Each cell in the table shown in figure 7 is the average percentage of customer journeys for that row (number of touchpoints) and that column (the baseline journey satisfaction rate). For example, the last entry on the bottom right shows that when we have 50 touchpoints, an average of 11% of customer journeys have baseline journey satisfaction rates between 96% and 100%.

The table can be visualized (figure 8) as a heat map with darker colors signifying higher numbers.

Figure 8: Heat map of mean % of customer journeys as a function of the number of touchpoints (rows 2 to 50) and the baseline journey satisfaction rate (columns 46%-50% to 96% to 100%)

The darker blue cells in figure 8 indicate higher average percentage of customer journeys in those cells. The lighter the blue the lower the mean percentage of customer journeys. The 81% that appears in the rightmost column of the first row in figure 7 is shaded the darkest blue in figure 8.

When there are 10 touchpoints or fewer, there are some squares that light up towards the left indicating that baseline journey satisfaction rates can dip into the 60% to 70% range. As the number of touchpoints increases, the colors get lighter but also more bunched at the right bottom of figure 8. This shows that baseline journey satisfaction rates now take on a number of different values but the percentage of customer journeys that have these values are not that high.

Conclusion


We distinguished three types of customer satisfaction rates -- touchpoint, baseline journey, and refined journey. The authors of the McKinsey article urge us to look beyond touchpoint satisfaction rates to refined journey satisfaction rates. They provide helpful information to improve these refined journey satisfaction rates.

But there's another difference that also merits attention -- the difference between touchpoint satisfaction rates and baseline journey satisfaction rates. The steps you take to improve the refined journey satisfaction rate will not improve the baseline journey satisfaction rate because the latter is driven purely by the mathematics of the siloed customer service set up. This awareness and some simple calculations give you the power to identify and improve those customer journeys that fall below the baseline journey satisfaction rates that your company can afford to tolerate.