The Customer Service Survey

Vocalabs' Blog


Customer Experience Isn't the Only Thing

Customer experience is an important part of how companies differentiate themselves competitively. But it's not the only thing. There are some circumstances where it makes sense for a company to not care about how its customers are treated. For example:

  • In some customers' minds, poor customer experience means low prices. Our favorite example these days is Spirit Airlines, a company which seems to delight in inflicting fees and inconvenience on its customers (though recently Spirit seems to be having a change of heart, possibly because of bad publicity). So if you are trying to stake out the market position of low-price leader, it may actually help you to make your customer experience worse.
  • If customers don't have a choice, then customer experience doesn't matter. Some of the lowest customer satisfaction scores are for cable TV companies, and it's not hard to see why: in most places you have only one real choice if you want cable (and increasingly, if you want high-speed Internet service), since the competitors don't have access to the physical infrastructure. So the cable incumbent doesn't need to invest in improving the customer experience to maintain its customer base.
  • When it's hard to switch to a competitor, it's possible to under-invest in the customer experience without losing too much business. Banks and mobile phone carriers are both industries where customers will often put up with terrible service because it's just too hard to change (or too expensive).

In all these cases, the company which provides a poor experience is relying on something else to keep customers coming: price, an effective monopoly, contractual commitments, etc.

But this can be a risky strategy. Subjecting customers to poor service builds up a reservoir of customer ill-will over time. If the market changes--or the company develops a bad enough reputation--it can be very expensive to repair the damage.

Pro Tip: Actually Listen to your Call Recordings Before Responding to a Complaint

Most companies of any size routinely record calls to customer service, and in some industries it's a legal requirement.

So when a customer complaints that customer service told him one thing but the company did another, it might be a good idea to actually listen to the recording and find out what the customer was actually told.

Otherwise you might wind up like Anthem Blue Cross, featured (in a bad way) in the LA Times. You might also wind up featured in a lawsuit.

The story is that an Anthem member needed a fairly expensive medical proceedure. Like a good and careful consumer, he called Anthem to make sure his doctor was in-network and the proceedure would be covered.

Upon getting a "yes" on both counts, he went ahead and got the treatment.

Experienced readers will know what happens next, because we've seen this movie before: After the treatment, Anthem denied the claim saying the doctor was out of network.

But this time there's a plot twist, since the patient had recorded his earlier call to Anthem. You know, the one where customer service assured him the doctor was in-network and the proceedure was covered. He appealed Anthem's decision and included a recording of the call with the appeal.

Incredibly, Anthem denied the appeal, claiming the member had never been told the doctor was in-network or the operation would be covered--points directly contradicted by the member's recording of the call.

This is about the point where the LA Times (and the lawsuit) come in. I assume that this will now be settled promptly in the patient's favor, since even faceless bureaucracies have their limits.

There are two morals to this story: For companies, do your homework and use the tools at your disposal. It can avoid an expensive and embarassing stuation.

For consumers, it's a really good idea to record calls to companies you do business with. After all, they're recording you, you should feel no hesitation to record them.

Combatting Survey Overload

I took a three-day trip this week to the CXPA conference in Atlanta, and was asked to take a survey about almost every single part of the trip.

If you're wondering why it's getting hard to get customers to respond to e-mail surveys, this is why. I was asked to take surveys by the airline, the hotel, the conference, and even the hamburger joint where I grabbed lunch after arriving.

Most of these surveys were quite lengthy (only the survey about my visit to the airline club had a reasonable number of questions), and more than one of them had over 75 questions. Do I even need to add the fact that the overwhelming majority of the questions were completely irrelevant to my personal experience? Or the fact that some of the questions were so badly written that I couldn't even figure out what they were asking?

All told, I was asked to answer somewhere between 200 and 300 questions about this short business trip. Most of the questions were irrelevant, and some were incomprehensible. I felt like my time had been wasted and my patience abused, all for an exercise which the companies clearly didn't care enough about to do properly.

And that's why it's getting harder and harder to get customers to respond to e-mail surveys.

So how do you do it right? Here's my advice:

  1. Keep it short and focused. For an online survey my mantra is "one page, no scrolling." If you can't fit the questions on a single screen, then your survey is too long. And while you may think there are 75 different things you need to ask about, the truth is that if you're trying to focus on 75 things then you are focused on nothing.
  2. Proofread, test, and test again. There is no excuse for bad questions, but an e-mail survey can go to thousands of customers before anyone spots the mistake. Have both a customer and an outside survey professional provide feedback (not someone who helped write the survey).
  3. Communicate to customers that you take the survey seriously and respect and value their feedback. Don't just say it, do it (following #1 and #2 above are a good start). Other things you should be doing include personally responding to customers who had complaints, and updating survey questions regularly to ask about current issues identified on the survey.

While these are all important things to do to build an effective online survey, the unfortunate truth is that the well is still being poisoned by the overwhelming number of bad online surveys. Customers have been trained to expect that any e-mail invitation to take a survey is likely to be more trouble than it's worth, and so I expect response rates to continue to go down.

More on Correlation

Just in case you need more convicing that correlation is not causation, spend some time browsing the Spurious Correlation Generator.

Every day it finds a new (but almost invariably bogus) statistical correlation between two unrelated data sets. Today, for example, we learn that U.S. spending on science and technolgy is very strongly correlated (0.992) with suicides by hanging.

In the past we've learned that the divorce rate in Maine is correlated to per-capita margarine consumption, and that the number of beehives is inversely correlated with arrests for juvenile pot possession.

These correlations are completely bogus, of course. The point is to illustrate the fact that if you look at enough different data points you will find lots of spurious statistical relationships. With computers and big data, it's trivially simple to generate thousands of correlations with very high statistical significance which also happen to be utterly meaningless.

Getting It Wrong

A little over a year ago, Forrester Research issued a report called 2013 Mobile Workforce Adoption Trends. In this report they did a survey of a few thousand IT workers worldwide and asked a bunch of questions about what kinds of gadgets they wanted. Based on those survey answers they tried to make some predictions about what sorts of gadgets people would be buying.

One of the much-hyped predictions was that worldwide about 200 million technology workers wanted a Microsoft Surface tablet.

Since then, Microsoft went on to sell a whopping 1-2 million tablets in the holiday selling season (the seasonal peak for tablet sales), capturing just a few percent of the market.

At first blush, one would be tempted to conclude that Forrester blew it.

Upon further reflection, it becomes clearer that Forrester blew it.

So what happened? With the strong disclaimer that I have not read the actual Forrester report (I'm not going to spend money to buy the full report), here are a few mistakes I think Forrester made:

  1. Forrester was motivated to generate attention-grabbing headlines. It worked, too. Ignoring the fact that there have been Windows-based tablets since 2002 and none of them set the world on fire, Forrester's seeming discovery of a vast unmet demand for Windows tablets generated a huge amount of publicity. Forrester might also have been trying to win business from Microsoft, creating a gigantic conflict of interest.
  2. Forrester oversold the conclusions. The survey (as far as I can tell) only asked IT workers what sort of tablet they would prefer to use, at a time when the Microsoft Surface had only recently entered the market and almost nobody had actually used one. That right there makes the extrapolation to "what people want to buy" highly suspect, since answers will be based more on marketing and brand name than the actual product. Furthermore, since this was a "global" survey, there was probably a substantial fraction of the population outside the U.S., Canada, and the E.U. who are unlikely to buy (or be issued) a tablet of any sort in the near future.
  3. Forrester let the hype cycle get carried away. I found many many articles quoting the "200 million Microsoft Surface Tablets" headline without any indication that Forrester did anything to tamp this down. Forrester's actual data basically said that about a third of IT workers surveyed said they would prefer a Microsoft-based tablet rather than Android, Apple, or some other brand, and if you believe there are about 600 million information workers worldwide (which Forrester apparently does), that's 200 million people. When that morphed into "Forrester predicts sales of 200 million Surface tablets," they did nothing to bring that back to reality.

All this is assuming that Forrester actually did the survey right, and they got a random sample, asked properly designed questions, and so forth.

At the end of the day, anyone who built a business plan and spent money on the assumption that Microsoft would sell 200 million Surface tablets any time in the next decade has probably realized by now that they made a huge mistake.

As the old saw goes, making predictions is hard, especially about the future.

Vocalabs Newsletter #78: Net Promoter and Customer Effort

We've published issue #78 of our newsletter. This issue has two articles: one about Net Promoter and Customer Effort, and the other about the importance of queue time in a customer service environment.

E-mail subscribers should have received their copies by now. As always, I hope you find our newsletter interesting and informative.

Correlation Is Not Causation

Anyone who works with statistics has heard the phrase "correlation is not causation."

What it means is that just because A is correlated with B you can't conclude that A causes B. It's also possible that B causes A, or that A and B are both caused by C, or that A and B are mutually dependent, or that it's all just a big coincidence.

Similarly, you can't assume that lack of correlation means lack of causation. Just because A isn't correlated with B doesn't mean that A does not cause B. It's possible that A causes B but with a time delay, or through some more complex relationship than the simple linear formula most correlation analysis assumes. It's also possible that B is caused by many different factors, including A, C, D, E, F, G, and the rest of the alphabet.

In reality, a linear correlation analyis mostly tells you the degree to which A and B are measuring the same thing. That's useful information but it doesn't necessarily tell you how to drive improvement in B.

I'm always a little disappointed when, in a business setting, someone does a linear correlation of a bunch of different variables against some key metric and then assumes that the things with the highest correlation coefficient are the ones to focus on. Correlation analysis can't actually tell you what causes the metric to go up or down: it's the wrong tool for the job. At best, it's a simple way to get a few hints about what might be worth a deeper look.

Actually understanding the drivers for a business metric requires a more sophisticated set of tools. A/B testing (where you actually perform an experiment) is the gold standard, but you can also learn a lot from natural experiments (taking advantage of events which normally happen in the course of business), and also from the basic exercise of formulating theories about what causes the metric to change and testing those theories against existing data. 

Issue #77 of Vocalabs' newsletter is published

We just published issue #77 of Quality Times, Vocalabs' newsletter. In this issue we have a pair of articles related to the design and interpretation of customer surveys. One is a few rules of thumb to follow when designing surveys; the other discusses how customers interpret satisfaction questions.

I hope you find this useful and informative, and welcome any comments and suggestions. 

Net Promoter and Customer Effort: Two Metrics Measuring Two Different Things

People often ask, "What's the right metric to use on a customer survey?"

The answer, of course, depends on what you're trying to measure. Often the survey has more than one goal, and this will require measuring more than one metric. Unfortunately, the people promoting the Net Promoter methodology have been promoting the idea that you only need to measure one thing (and, of course, that one thing is their metric).

As a case in point, we have a client currently asking both a recommendation question (similar to Net Promoter) and a customer effort question. Customer Effort is a metric designed to measure the roadblocks a customer experiences in trying to get service, and it's a good way to gauge how smoothly a particular transaction went. Net Promoter, in contrast, measures a customer's overall relationship with the brand and company.

In this survey we noticed a curious thing: a meaningful percentage of customers who both said they would recommend the company, but who also said they had to go through a lot of effort to get what they wanted on the customer service call.

This should be surprising to anyone using Net Promoter to measure a particular customer experience--the theory being that customers who just had a bad experience will be less likely to recommend the company.

That theory may have some truth on average, but when it comes to individual customers there's clearly something else going on.

So we listened to a number of the interview recordings to better understand what the customers were saying. And the message was loud and clear: These customers had a bad customer service experience, but were loyal to the company for completely unrelated reasons.

The recommendation question was doing exactly what it was supposed to do: measure the customer's overall relationship with the company. And the customer effort question was also doing exactly what it was supposed to do: find the ways the company made it hard for customers to get the service they expected.

The lesson is simple, but often needs to be repeated. Ask the question about what you want to know. Don't expect a survey question designed to tell you one thing to measure something else.

Net Promoter and Customer Effort are two different questions which measure two different things.

How long did you wait?

One of the oldest complaints about customer service is having to wait on hold to talk to a person. It's still a problem from time to time in many companies, and we published some research on hold times as part of the mid-2013 NCSS Banking report (see page 3 of the PDF report).

We had a recent opportunity with a client to explore how well customers estimate their wait on hold. Anecdotally, we all know the customer who said he waited ten minutes but only actually spend 30 seconds in queue. For this client, they were able to supply us the actual time in queue for each customer who completed a survey, which we compared to the customer's estimate of the wait for an agent.

The results were interesting and surprising. It turns out that an individual customer's estimate of the time spent waiting bears almost no relationship to the actual queue time for that customer. There were plenty of instances of dramatic over- and under-estimates of the wait time. I'm talking about people who claimed they had to wait ten minutes but actually spent less than a minute in queue--or, conversely, people who said it was under a minute when it was actually several.

However, on average, customers' estimates of the wait time were astonishingly accurate. For example, taking all the people who said their wait time was "about two minutes", and averaging their actual queue time, it was surprisingly close to 120 seconds.

We also found that both actual and perceived wait time correlated to IVR and call satisfaction, but the perceived wait time was a stronger relationship. I suspect this may have to do with the customer's emotional state: the more annoyed he is with the call, the less satisfied, and the longer he thinks he had to wait to speak to someone.

Finally, there's a significant minority of customers (I'm guessing around 20%) who apparently are including the time spent navigating the IVR in their estimates of the wait to speak to someone. So even if the actual queue time was short, a long and complicated IVR makes some people feel like they're waiting for an agent.

So the lessons we learned are:

  • Queue time still matters in customer service. It feels a little old-school in this age of social media and natural language IVR, but make sure you're answering the phones promptly.
  • The actual queue time and what the customer thought the queue time was are two different things. You're probably measuring the former but not the latter, but it's the customer's perception which counts.
  • Making customers go through extra steps to reach a person makes them think it took longer to reach someone, and makes customers less satisfied with the service.

We're Watching You, Comcast!


Comcast and Time Warner have launched a PR offensive to try to convice people that it's going to improve it's customer service in advance of their pending merger, as evidenced by a pair of puff-pieces in USA Today and Marketwatch today.

Comcast, of course, is the company which was far behind its peers for customer service in the recent National Customer Service Survey results. Time Warner did better than Comcast, but is still below most of the others.

Speaking as a Comcast customer myself, I truly hope the company is mending its ways in customer service. But I'm also very skeptical. It takes more than good intentions and noise from the executive suite to make this kind of change: it requires changing the way thousands of individual employees interact with customers on a daily basis, it requires fixing broken processes which prevent resolution of customer issues, and most of all it requires time and hard work.

Many customer service initiatives fail because, while the leadership is willing to talk a good game, they aren't willing to devote the effort and resources.

Fortunately, though, we won't have to take Comcast's word on whether their customer service is improving. We will see soon enough, through the ongoing customer feedback in the National Customer Service survey, whether they are actually making any improvements. I look forward to seeing the results over the coming months.

So Comcast, it's great that you're talking about improving service. But we're watching you.

Customers Don't Give Points, They Take Them Away

What does "Very Satisfied" mean? Does it mean "Outstanding job, above and beyond expectations?" or does it mean "I don't have any complaints?"

Many people who receive customer feedback think it means the former. But in most cases, the data suggests that it actually means the latter. In other words, if a customer gives you the top score in a survey, often times it just means you did your job.

Case in point: for one of our clients, we are following up on one of the core satisfaction questions by asking the customer to explain the reason for his or her rating. Because this is an interview format, we are getting a response over 90% of the time.

When the customer gave the top rating, "Very Satisfied," 99% of the reasons given are positive (and the small number of negative comments were mostly about unrelated things). This isn't surprising.

But when the customer gave anything other than that top score, even the mostly-OK-sounding "Somewhat Satisfied," 96% of the reasons the customers gave for their rating are negative.

In other words: If the customer didn't give the best possible score, there was almost always a specific complaint.

We see a similar pattern in most questions where we ask the customer to rate the company or the interaction. Another client which is using a 0-10 point "recommendation" question (aka "Net Promoter"), we see over half the people who gave an 8 out of 10 had some specific complaint (and nearly everyone who gave 6 or below had a complaint).

The notion that the middle point on the scale is somehow "neutral" (even if we call it "Neutral") is simply not consistent with how people really answer these kinds of questions.

Instead, most people start near the top of the scale and mark you down for specific reasons. If the customer has nothing to complain about, you get something at or near the best possible score.

So in most cases, customers don't give you a better rating for better service and a worse rating for worse service. Instead, they give you a good rating and take away points for things you did wrong.

Working Backwards to the Technology

Via Daring Fireball, I found an amazing video of Steve Jobs from 1997 talking about his philosophy of business. This was after he returned to Apple, but before any of the iProducts (iMac, iPod, iPhone, iPad) which would make Apple what it is today.

The whole thing is worth watching, but the core lesson is captured in this quote:’ve got to start with the customer experience and work backwards to the technology. You can’t start with the technology and try to figure out where you’re going to try to sell it.

The remarkable thing about this quote (other than the fact that he used "customer experience" over a decade before it became a hot buzzphrase) is that this is exactly backwards from the way almost every other technology company develops its products.

You could substitute the word "capabilities" for "technology," and Jobs' lesson would still be true, and it would still be backwards from the way many other companies develop their products and services.

Most companies (even startups) begin product or service development with the capabilities they have (or could quickly acquire), and try to find ways to attract customers by packaging those capabilities up at an attractive price. That favors inertia over developing a good customer experience. It can also lead to feature clutter, since the temptation is always there to include something "because we can."

The Apple approach, on the other hand, puts the desired end result front and center. Along the way, some compromises will inevitably have to be made where there are things which can't be done at a reasonable price. But by starting with the customer experience and working backwards, you keep the experience front-and-center during the whole development process.

So why don't more companies follow the experience-first strategy? My sense is that it's just hard. It's hard to stay focused on the end product, it's hard to say "no" to cool capabilities which don't enhance the overall customer experience, and it's hard to accept that a longer development cycle may be required if a critical piece of the puzzle doesn't exist yet.

This Survey IS a Test

Last week, Bank of America invited me to take a customer survey.

The survey was nothing special, but at the end it gave me this message (click to embiggen):

In case you can't view the image, at the end of the survey it displayed:

Elapsed time: 366
Pass/Fail: Pass

I was unaware that my opinions about Bank of America were being graded. But as long as they were, I'm glad to see that I passed.

However, the more competitive side of me wants to know what my letter grade would have been, and whether this was graded on a curve.

Who Thought This was a Good Idea?

Via Consumerist comes the story of a guy named Guy. Guy shops at Staples, and is a member of Staples' reward program, and Staples recently invited Guy to participate in a market research survey. In exchange, Guy would get a $5 check for his trouble.

Instead of a check, though, at the end of the survey Guy got this message:

Your opinions are extremely important to us. Unfortunately, we have reached the target number of completes from your group today. However, your time and efforts are greatly appreciated.

Thanks again for your support and participation.

Apparently, sometime between when Guy started the survey and when he finished it, Staples filled its quota of responses. And rather than spend one penny more than necessary on survey incentives, they gave Guy the "Sorry, Sucka!" message.

The survey invitation claims that the survey was being hosted by an "independent research company," but any market research company engaged in this sort of amateurish behavior deserves to be drummed out of the Survey Corps.

I get that surveys sometimes go over quota, but any responsible professional will plan and budget for that, rather than risk angering the very people who are helping you collect your data. And in the bigger picture, five bucks (even five bucks times a hundred) is chump change.

In the end, when reporters from Consumerist asked Staples for their comment, Staples did the right thing and sent Guy his five bucks. But you have to wonder how many other Staples customers got the same message and decided to seethe quietly rather than complain to the media.

I'm guessing those customers are going to cost Staples a lot more than five bucks in the long run.

Vocalabs Newsletter Published: NCSS 2013 Data

We just published Issue 76 of our newsletter, Quality Times. In this issue we announce the availability of 2013 survey data for the National Customer Service Survey in both Communications Services (formerly Mobile Phones) and Banking. We are adding five new companies to the Communcations Services report: CenturyLink, Comcast, DirecTV, Dish Networks, and Time Warner Cable.

As always, I hope you find this useful and informative, and welcome any comments and suggestions.

New NCSS Reports Available


We just published updated NCSS reports, including 2013 data for both banking and communications companies.

We've added coverage of five new companies in the communications sector, bringing it to nine companies: AT&T, CenturyLink, Comcast, DirecTV, Dish Network, Sprint, T-Mobile, Time Warner Cable, and Verizon.

The banking report covers Bank of America, Chase, Citi, and Wells Fargo.

In addition to the latest reports, you can still download our older research if you want to view earlier reports.

Well-Oiled Company = Well-Oiled Experience?

I've worked with a lot of client companies over the years at Vocalabs. Some companies are easy to work with: they are well-organized, people are focused on getting the job done, decisions are made easily, and the lines of communication work well.

Other companies are not so much fun to work with, for a variety of reasons. Internal politics, disorganization, people more worried about keeping their jobs than doing their jobs, muddy lines of authority, and disspirited employees, to name a few problems.

As a consumer, some companies are much easier and more enjoyable to do business with than others. And given the choice, I would rather take my business to a company which provides the more positive experience.

I've noticed over the years that the companies I enjoy doing business with as a consumer also tend to be easier to work with professionally. That is to say, if I like buying from the company, I'll probably find that the company is organized and efficient  when it comes to hiring Vocalabs to work on a project.

The opposite also seems to be true: If the company is disorganized internally, chances are I probably won't like doing business with them as a consumer. On the other hand, just because I don't like buying from a company it doens't always follow that they're going to be hard to do business with.

It seems that having a well-oiled corporate machine is a necessary requirement for having a good customer experience. That makes sense, since a disorganized company almost can't help but screw things up as far as the customer is concerned. Many common customer complaints (mistakes, inconsistency, indifferent service, lost records, etc.) are the inevitable outcome when a company can't get all its internal resources lined up and marching in the same direction.

It's not true, however, that an organized and efficient company will always have a good customer experience. The company may be well-organized, but simply not focused on the customer. In those cases, though, the company often knows it's competitive edge isn't coming from the customer experience (maybe it's the low-cost leader instead), and there's a deliberate choice to not invest to improve the experience.

So an important prerequisite to delivering a positive customer experience is having a company which generally runs smoothly. But that, by itself, won't lead to a great customer experience. The company must also deliberately choose to go down the path of customer experience.

Vocalabs Newsletter #75: Customer Experience Journey Map of Christmas

Somehow, a confidential e-mail from the North Pole was recently misdirected to me. I thought the contents would be interesting to readers of this newsletter, so I am taking the risk of coal in my stocking to reveal its contents.
You can read the e-mail in the most recent edition of our newsletter, Quality Times. As always, I hope you find it interesting and informative.

Fooling the Customer

Time had an amusing article this week about The Telemarketing Robot who Denies She's a Robot (complete with recordings!). A company offering health insurance had, for a time, a phone number answered by an a cheerful-sounding woman who would ask several questions about the caller's insurance and respond to conversational questions.

Ask if she was a robot, and she would laugh and insist that she was a real person.

Except that she wasn't, as would quickly become apparent by the awkward pauses before she would reply, her clearly limited set of responses to questions, and the way she would speak the exact same phrase (words, tone, timing) to different callers. Most likely, "Samantha" was a set of prerecorded messages triggered by someone listening to the caller and selecting the best response from a list. This would allow the use of really cheap overseas employees with poor English skills.

In the speech recognition industry about ten years ago there was a debate about how "human" an automated system should be. On the one hand, some designers believed that the best speech systems were the most natural and conversational--in the ideal world, you would call your bank and never know whether you were talking to a person or a machine.

The other view--which I hold along with people like Bruce Ballentine (author of It's Better to be a Good Machine than a Bad Person)--is that you should always make it clear to the caller whether they're interacting with a machine or a person. Leaving aside the fact that the technology is nowhere near advanced enough to allow for a true conversational experience, people just don't like to be fooled.

People care very deeply whether they're talking to a machine or not. It's not that talking to a machine is bad (witness the willingness to use Apple's Siri service). It's because talking to a person carries social context and talking to a machine doesn't, and you can interact with a machine in ways you wouldn't interact with another human.

We observe this in people's willingness to play with the machine and explore its capabilities. While sometimes people will observe social norms when talking to a computer (for example, saying "Please" and "Thank you"), they also feel free to break outside the box. The recordings of the Time reporters taunting "Samantha" are a great example (they can't get her to repeat the exact phrase "I am not a robot"). Ballentine called this the problem of the "Monkey-butt user," after someone in a usability test who randomly said "Monkey-butt" to the computer to see how it would react.

(As an aside: Bruce's book is the only one I know which has "Monkey-Butt" in the index. Look it up yourself.)

It turns out that humans care so much about whether we're talking to a machine or not that, if we suspect a machine it trying to fool us we will spontaneously begin a series of Turing tests to find out if it's really a machine or a person. Current technology has a long way to go before it can get past someone who is determined to discover the truth.

So the lesson is simply this: Don't try to fool your customers. It won't work, and they won't like it.

Most Depressing Survey Ever

From Tumblr user pupismyname.

The Limits of Metrics

Andy Beaumont recently wrote an article, The Value of Content, which is (at its core) a screed against the annoying web design trend of making readers click through overlays in order to get to the actual page they want to read. Beaumont struck a nerve with his "Tab Closed; Didn't Read" collection of annoying examples.

Beaumont's article is an excellent argument for all the reasons against this technique, and one point in particular struck me. In responding to the argument that overlays get used because they work, he writes:

This is what happens when analytics make decisions for you...Analytics will tell you that you got more “conversions”. Analytics will show you rising graphs and bigger numbers. You will show these to your boss or your client. They will falsely conclude that people love these modal overlays.

But they don’t. Nobody likes them. Conversions are not people. If you want the whole story here you should also be sat in a room testing this modal overlay with real people. Ask them questions:

  • “Do you like that overlay asking you to sign up for the newsletter?”
  • “Do you understand what will happen if you do sign up for it?”
  • “Do you know that there is content behind it?”
  • “Do you know how to close it to get to the content?”.

This gets to the heart of the difference between customer feedback and other metrics and analytics: there are two sides to every story, and if you're not collecting customer feedback then you're only getting one side of the story. In this case, the overlay may be effective at getting newsletter signups, Facebook likes, etc., but not at all effective in generating actual customer engagement or useful sales leads.

The key is to recognize that the signups and likes are not, in themselves, the true business goal. The true business goal is the customer engagement or the sales lead. And it's important to recognize that the reason behind the signups and the likes is more important than the actual signup or like.

Put another way, if someone signs up for your newsletter because they want to receive your newsletter, there's a good chance that's a useful sales lead. But if someone signs up for your newsletter because they thought it was the only way to close an annoying window and get to the article they wanted to read, that's just an annoyed web surfer.

We see this same thing play out all the time in the customer service world when companies get overly focused on a particular set of metrics and forget to ask what's really going on for the customer. The classic is using containment to measure how well an automated customer service system is working, and assuming that anyone who hangs up successfully self-served.

We even see this problem when it comes to customer feedback surveys. Many companies track their survey scores and assume that if the number is going up things are getting better, and vice-versa. But it's also possible that employees are gaming the survey, some customers are being blocked from taking the survey, or customers' expectations are changing.

The lesson here is that any given metric is, at best, an approximation for the real business goal. It's important to always keep that in mind, and constantly ask not just "what" is happening but also "why." And the most powerful tool for doing this is using a truly closed-loop process, where you close the loop with the customer, the business, and also your metrics and feedback.

Comcast CEO: Delusional or Just Spinning?


Brian Roberts, Comcast's CEO, was interviewed yesterday for the radio show Marketplace. The company has a reputation for poor customer service, and so at one point the host asked Roberts to respond.

His answer to why Comcast has such a bad rap: "...we have about...350 million interactions with customers a year, between phone calls and truck rolls....You get one-tenth of one percent bad experience, that's a lot of people. Unacceptable. We have to be the best service provider or in the end, this company won't be what I want it to be."

(You can listen to the full interview on the Marketplace website, without my edits for brevity.)

So his explanation for Comcast's reputation is that they have so many customers that there's always going to be some tiny fraction who get bad service.

Actually....that's not quite what he said. What he said was that if they had one customer in a thousand get bad service it would be a lot of people. He sort of implied that's the level Comcast is operating at, without actually saying it.

Which is good in a way, because if Roberts actually believes that only one in a thousand Comcast customer interactions is a bad experience, this represents an astonishing degree of executive delusion, bordering on clinical madness. Having 999 of every 1,000 customer experiences be positive represents world-class customer service, at a level probably not attainable by any company with millions of customers.

So how is Comcast actually doing? It so happens that we have some data on that topic, since we have been interviewing Comcast customers since earlier this year as part of the National Customer Service Survey. These interviews happen a few minutes after a customer calls Comcast, and can be fairly extensive. We published some preliminary results a few months ago which showed that (based on very early data with a big margin of error) Comcast's customer satisfaction with its customer service is well below industry peers.

And those peers are not companies famous for good service, either: AT&T, Verizon, Sprint, and T-Mobile. And this isn't just a survey about reputation, like a JD Powers. We are asking about a specific call to Comcast which happened immediately before the interview.

I have to assume that the CEO of Comcast has at least some idea that his company really doesn't provide good service. But maybe not. It's possible this is another example of the corporate Dunning-Kruger Effect. But more likely, he's been well-trained by media handlers to spin the question really effectively.

But whether he's delusional or just offering his spin, it's disappointing that Roberts can't seem to acknowledge what is apparent to so many people outside the company. Because the first step to improving your service is to admit that it needs improvement.

Issue #74 of Vocalabs Newsletter

We just published Issue #74 of Vocalabs' newsletter, Quality Times. We're calling this The Listicle Issue, since we are featuring two "Listicles" on developing a good customer feedback process: 8 Rules for Writing a Not-Awful Survey, and 5 Things Your Survey Should Trigger.

As always, I hope you find this interesting and informative. I welcome any comments or suggestions.

8 Rules for Writing a Not-Awful Survey

Writing a good survey isn't hard, but there are some gotchas if you've never done it before. Here are a few rules of thumb to help avoid the biggest mistakes:

  1. Keep the survey short:
    • Phone interviews should be under 5 minutes
    • Online surveys should fit on a single screen without scrolling
    • IVR surveys should be 5 questions or fewer
  2. Keep the questions short and use simple language. Avoid jargon or brand names, since there's a good chance customers won't recognize them.
  3. Always begin by asking the customer to rate the company as a whole, even if that's not what the survey is about. This gives customers who have a problem with the company a chance to get it off their chest so they won't penalize the representative.
  4. Put the most important questions (usually your tracking metrics) near the beginning. That way they are less likely to be biased by other questions and more likely to be answered.
  5. Be as consistent as possible with your rating scale. For example, don't switch from a 0-10 scale to a 1-5 scale.
  6. In the U.S., it's conventional for higher numbers to be better. Don't make "1" best and "10" worst as it's likely to confuse people. (This rule may differ in other cultures).
  7. Always have at least one free response question.
  8. Plan on making regular changes to the survey. You won't get it perfect the first try.

Following these rules won't necessarily give you a great survey, but breaking these rules will almost always make it worse.

Syndicate content