Bad Customer Service—Why Do It?

Megan McArdle over at Asymmetrical Information kvetched about her horrible customer service experience.  I haven’t had any of my Sony products break down, but then again, outside of one computer and one monitor that we bought 11 years ago (the monitor still works; the computer had been gutted entirely, so only the case—which wasn’t all that good—remained), my family doesn’t do much in the way of buying Sony products.  I can sympathize with her, however, because bad customer support experience seems to be a constant.  Let’s try to come up with some compelling hypotheses regarding why this is so (note that I stole some from the comments section, but I forget who said what when, so sadly, the people who came up with some of these ideas won’t get credit).  In numerated form!

1.  Most individuals never use customer support lines, so the portion of the customer base doing so is fairly limited.
2. Customer support, particularly first-tier, is a menial job, so most of the people hired for lower-level work lack the knowledge necessary to solve problems.  Because of this, they are giving little leeway and do not get paid much.  In part because of those, they quickly become bitter as screaming customers demand everything and this job just isn’t worth it, leading to poor first-tier behavior.
3. Customer support is a cost center, meaning that it has no direct revenue sources, so it is difficult to argue for expansion and improvement of the process at the expense of revenue-generating processes.  This leads to a situation like hypothesis #2, or…
4. When customer service functions are offshored, they generally go to firms which specialize in customer service functions rather than specific problem-solving functions.  Bob’s Customer Support Shack hires people to answer phones for many different companies, products, and fields, meaning that no first-tier support person actually has the knowledge and training to solve problems outside of reading a script.
5. (From Tyler Cowen) People are geared toward getting a solution once they call technical support, so if the solution does not work, they become relatively more disappointed than they otherwise would in other situations.
6. (From Tyler Cowen) Uncertainty regarding when a problem will be fixed is more painful than knowing a fixed date, even if that date is relatively far in the future.
7. People don’t really care about customer support very much, at least compared to other aspects such as price, qualiy of the good, features compared to other products, and aesthetics.  Thus, a company is better off focusing on the product at the expense of customer support.
8. Figuring out what good customer support is turns out to be difficult and there are inherent tradeoffs.  For example, having a person stick to your problem until he fixes it sounds like good customer support…but if all of the staff is doing this, other customers likely to have long waiting times on the phone, which would make customer service bad for them.
9. Determining quality with customer support is difficult.  There are few reliable metrics associated with the process.  Sometimes a very satisfied customer will write a letter, but most of the time, the feedback loops are shaky.  Customers generally do not provide accurate judgments of how their service was (you know all of those cards that ask you to rate customer service at places?  How often do you fill one of those out?  That’s right:  practically never), so management is left looking at other benchmarks, like average wait time…which isn’t all that good of a support measure.
10. Customers want too much and do not understand the basic tradeoffs involved.  I don’t want to pay extra for a product, but I want excellent support.  I want people who can solve my problems, but I don’t want to wait for it, so they better have a large enough staff of highly-trained individuals even though I’m not willing to pay extra for that.

So, how do these hypotheses look?  #1 has some merit, but I’m not sure that it’s all that important.  Alienating a decent fraction of your customer base—even if it’s only 5%—does not seem like a sound business strategy.  Numbers 2-4 seem pretty much on the ball.  First-tier support personnel, especially in IT-related fields, generally are not experienced and basically act as call screeners, filtering out the easiest problems.  I don’t really like Tyler Cowen’s hypotheses that much.  Yeah, I am often expecting a solution when I call, but I don’t believe that this really biases my thinking so much as to make me believe that support is worse than it really is.  Impatience regarding the uncertain is a stronger answer, but again, it only describes a relatively small subset of customer support problems.

I’m not sure about #7.  A lot of people do purchase warranty extensions, customer support plans, in-home technician service, etc.  This would apply to #10 as well.  There are some people who are really begging for a free lunch, but most people understand the nature of tradeoffs, and looking for the best deal or desiring a perfect situation doesn’t really diminish that.  #8 has some merit, although a lot of the tradeoff has to do with hypotheses 2-4, in that a well-staffed and well-trained customer support staff with a lot of leeway will eliminate most of the tradeoffs involved.  #9 leads to some bad management strategies and pressure to keep time per call down, but I don’t know if it’s really that powerful of a hypothesis.

So basically, 2-4 seem to be the driving hypotheses, with #8 tagging along.  I’m not sure how much (if any) role the others play.