Customer support: an interview with Joel Spolsky

Gerry Gaffney Service design 1 Comment

Download (mp3: 10.6MB, 23:10) Good customer support practices

Share this episode



Transcript

Gerry Gaffney:

Joel Spolsky runs Fog Creek software, who make FogBugz bug tracking software. He’s also written User Interface Design for Programmers and Joel on Software. In between, he maintains the Joel on Software blog.

Joel, welcome to the User Experience Podcast.

Joel Spolsky:

Thank you for having me.

Gerry:

Recently, you wrote an article entitled Seven Steps To Remarkable Customer Service
and step 1, the one that caught my eye said “Fix everything two ways”. Can you tell me about that theory?

Joel:

Yeah, well, one of the things that is the most frustrating about customer service is when you call up and there’s something that the organisation you’re calling is clearly doing wrong, and if they fixed it, they would be much happier, and you would be happier too, but they would be much happier, and so you call them and say; “You guys really need to fix such and such and then you won’t get so many damn phone calls.” And then it turns out that they don’t have any way to do that, they don’t really believe in fixing things, or you’re talking to some kind of outsourced customer service provider that has no way of actually talking to the company itself.

So this was brought to a head once when I bought a printer from Hewlett Packard that had a bug in it that would have been fairly easy to fix, that caused us to spend several hours getting this printer up and running. It was just a generic laser printer, running on generic Windows XP and there was no reason it should have taken so long to get set up, and had so many incompatibilities and problems. And so when I called them, I said; “You know, I got it working,” and the tech support people kept wanting to help me get it working and I said; “That’s not what I need, I need you to change this little tiny thing in the way that your setup program works, so that nobody else has this problem.” And I felt like I was being rather altruistic and they didn’t really care. [Laughter.]

They had no interest whatsoever in improving their process. So we’ve always done the opposite, and one of the things that it’s allowed us to do is grow revenues by tenfold without increasing the number of people that it takes to provide tech support. My feeling is that we would have ten times as many tech support people right now if we hadn’t practised this fixing everything two ways.

Gerry:

So, when you’re talk about two ways you mean fix the immediate problem for the customer and then fix the underlying cause, whatever that may be.

Joel:

Yeah, whatever it may be that causes the thing to happen. And it may not be something that’s broken – it may be something that a customer is misunderstanding because you haven’t explained it well enough. It may be something that the customer is understanding perfectly well but they’re not happy about, and then you need to fix that. So we sort of have an obsession of… literally whenever we see any kind of request or question, asking ourselves what can we do so that we never get this question ever again.

Gerry:

That implies that your development staff and your support staff are very tightly coupled.

Joel:

Yes, and we are luckily all in one place here, and that is, I believe, one of the benefits that we get of not outsourcing customer service, doing customer service in our case in a very expensive place with far over-educated highly technical people – office space in the middle of New York City – and it’s costing me probably six times as much per person doing tech support as it would cost to find the lowest cost provider in Bangalore or Hyderabad.

But as a result, if they have a problem they can’t solve, they go talk to a developer across the hall and every day at lunch they see the developers. So when there are interesting problems that they think there’s a way we could prevent them from happening, they can talk to the developers at the lunch table or in the kitchen.

Gerry:

That’s interesting because many organisations treat the tech support as something to be outsourced as cheaply as possible, and they look at it purely on a dollar per call metric, but you’re implying that that’s not a very sensible way to do things.

Joel:

No. We look at it on a dollar metric, not a dollar per call – a total dollars to the organisation – metric. And so I feel like what we’re doing is much more cost-effective. Like I say, if we hadn’t been doing this, I think we probably would need about eight people to do tech support today, and as it is we need one, and another person to do sales.

Gerry:

Really, one person doing tech support? That’s pretty impressive. What about when you scale up, though, is that going to work for someone like a large telco that by definition is going to have many many many incoming calls?

Joel:

That’s an interesting question… I have never run customer service operations for a large telco, so I’m not necessarily the right person to ask. On the other hand, to an organisation like that, every incoming call is a chance to make a customer fall in love with your company, not that they ever do that.

Gerry:

Or a chance to lose them forever.

Joel:

Or a chance to lose them forever. And so, when I think about, if I were in charge of marketing for, let’s say a large airline… Marketing and customer service are different silos in a large airline. Customer service people are trying to reduce their costs and so they’ll install interactive voice response systems for everything common and you’ll call them up and they’ll say if you want to know about arrivals and departures press 1. If you’re in charge of marketing, all you’re doing is trying to figure out how to reach potential customers and customers and get messages to them, while this other department in your company is doing everything possible not to talk to customers. So if those were somehow in the same silo, you might say… If we have, let’s say that you’re in an airline, obviously you have peak times and non-peak times and at peak times everybody’s very busy answering the phone, but at non-peak times there’s probably a bunch of people sitting around. Have them pick up the phone and talk to some of those people that are just calling to enquire about an arrival or departure and pick them up and be friendly to them and say; “Have a nice day,” and; “Can I help you with any future flights?” maybe or; “Do you know about Continental Airlines Mastercard?” and impress them with the quality of your service. So one of the things that we do here, we make an attempt to do here – you didn’t notice it because you reached us after our office hours, but when you call Fog Creek you get a human. And that’s the only way we’re ever going to win hearts and souls is by actually talking to customers and impressing them that we’re friendly and we’re here.

Gerry:

Yeah, it’s quite interesting, I spend a lot of time out with various users on site and one of the things they always comment about is organisations, when they do ring up when they contact a human, it’s always notable to them that that’s the case.

Many of us of course have a very low expectation of what will happen when we ring customer support and I guess some people could say that we blame it on the fact that typically we’re dealing with technical people who by definition, you know, don’t know how to interact with other humans! But you’re fairly technical and you espouse excellence, technical excellence and customer support. How is that?

Joel:

Hmm. I don’t think that the problem is necessarily that you’re reaching technical people who don’t know how to interact with humans – I think actually it’s quite the opposite. Anybody who can… I mean there’s this sort of a major problem that people who are able to do good customer service, whether it’s tech support, inside sales, in a highly technical organisation are not really going to enjoy doing it for a career. And after six months to a year, they’re going to get extremely bored and need to move on, and so one of the biggest dilemmas we faced was how do you provide tech support when the only people that can do tech support are so skilled that they could be doing a job that is better than tech support? And the way most people deal with this is… you know, one way to deal with this is to hire under-qualified workers, and that’s the most common. Another way to deal with it is to try to outsource it to countries where the qualified workers may not have as many opportunities. And you know, for a while a lot of things were outsourced to India under that rubric. But actually now people in India have much better opportunities I think than doing tech support for American companies, so that’s not working so well.

For us, the solution that we found was making it a part of a career path. In our particular case, tech support, inside sales, QA, testing and marketing/communications are all on the same career path in a management training program that we’ve set up. And so when we hire a tech support worker they’re actually coming in and doing a management training program that involves rotations through tech support, inside sales, marketing, testing, QA, beta management, product management, but also includes night and weekend classes that lead to a Master’s degree, sort of an MBA equivalent in Technology Management. It costs us a bunch more, but we have extremely good people doing those things.

Gerry:

And you still reckon from a bottom line perspective it’s cheaper than outsourcing?

Joel:

Oh, absolutely. Because even though an individual call may cost three or four times as much, they can solve the problems much faster, because they’re far more qualified. … A lot of times if you have a FogBugz problem, when you call tech support and you try to explain your problem you’re talking to someone who has experience programming, and they will often go into the code with you, debug it, and get you up and running in 10 minutes. Whereas a traditional tech support organisation would spend at least an hour and a half discovering that they just do not have the skills to debug this problem for the customer. So even though theoretically on an hour-by-hour basis it may appear to cost more, we believe that it’s a very economical way to do tech support.

Gerry:

Ok. Now, to change tack slightly… In, I think it was in your blog you described a fantastic customer experience you had with an organisation called Lands’ End. You had mistakenly ordered a batch of shirts with the wrong coloured lettering or something and they replaced them for you at no cost to you whatsoever. Which is a great customer experience but you could sort of argue that’s okay for a quirky little clothing manufacturer or a quirky little software manufacturer but it wouldn’t work in the, let’s say in the real world of banking or telecommunications or the like.

Joel:

Well, maybe not in Australia, but Lands’ End is huge, so they’re not quirky and they’re not little.

Gerry:

[Laughs.] Ok, I apologise unreservedly to Lands’ End.

Joel:

Yeah, the only quirky thing about them is that they answer the phone before it rings, it’s very unusual to call them, because they must have a system, I’m not sure about this, but they must have a system where the next customer service agent is told to stand by because they’re getting the next call, even in advance of that call coming in, because the minute you finish dialling they’re sitting there saying; “Hello, Lands’ End, have a nice day, how may I help you, etc” and there’s just this sort of this extreme bend over backwards to make you happy. And an interesting point is that the shirts are not the cheapest, and the quality… the quality is good, but the shirts, I mean you’re ordering shirts that cost $20, $30 that you could probably find for $12, and the cost to them is about $6. So even though they made the shirts wrong and then completely replaced them they probably didn’t really lose money on me in that particular case. So I think there’s sort of a difference between a high road service, a high road company that basically wants to provide the premium service at a fair but not cheap price with, you know, extremely good customer service and very very good support, versus the low cost providers that are always trying to do everything on the cheap. And the problem happens when consumers want to use a low cost provider and expect to get premium service at the same time. And that’s when they’re disappointed in the experience they get. You know, well of course you know there was nobody to call and they didn’t answer the phone, you’re only paying $4. People get upset about the poor service they receive from domain name registrars like godaddy.com where they’re paying six bucks! You know, how can they, think about it, like [laughs], how could they even take your call if you’re only paying them $6?

Gerry:

I guess it’s an interesting trade-off between those price point, those things that are sold on price and those things that are sold on quality and, you know, do people really pay for quality or are people willing to pay for that premium in general.

Joel:

Yeah, and there’s room for both of them. As a business person I vastly prefer to take the high road and to be providing the premium product rather then always trying to figure out how I can be beaten on price… I’d much rather be the fancy High Street store, the Neiman Marcus that sells expensive products to rich ladies, than be the Walmart of the world, that sells everything at the lowest possible, cheapest possible price. Simply because the profit margins are greater, and because there’s a lot more opportunities to make a mistake and because if you’re going to be the Walmart or the cheap provider of the world you can’t afford a very nice chair to sit in all day, you’ve got to buy a cheap chair, so you’re always trying to lower your costs.

Gerry:

Why do so many organisations get it so wrong? And I guess I think in terms of telcos. Whenever I’ve got a serious problem with my telco, which is not, you know, the cheapest telco in the world, I have to ring them up and threaten to switch carriers before they’ll actually do anything.

Joel:

Well you’re lucky you have the ability to switch carriers. [Laughs.]

Gerry:

Yeah, I guess.

Joel:

Not everybody is really in that position. And that’s actually the source of the problem really, is that a lot of the telcos may have grown up as monopolies and you know like they say; “We’re the telephone company, we don’t care because we don’t have to”. I can’t really explain it. Part of the problem is that as soon as they start marching down the road towards reducing costs as much as possible, partially because they’re trying to make a profit, but even more because highly competitive businesses, a lot of customers are very very much driven by what’s the lowest cost provider, and so they’ll do all kinds of things, often very short-term things that don’t pay off, to try and reduce those costs, even if it generates enormous amounts of customer churn. An interesting thing, at least in the United States, there are at least three major cell phone providers right now that are remaining. There’s Cingular (the new AT&T) is one, Verizon and Sprint are probably the last three standing, a bunch of small regional ones. But those are the major players and they all have horrible service, and there’s massive churn, people leave them all the time, and as a result, you know, it sometimes seems like the only advertisements you see on television are for cell phone companies, because they have so much churn and so many people are leaving them because they have bad service that they need to advertise like crazy to just keep enough new customers coming in to make up for the ones that they just lost.

And it seems to me like any one of those companies could wake up one morning and say, you know what, let’s stop advertising completely and just figure out how to keep our existing customers and if we can be the one cell phone provider that keeps customers rather than losing them, eventually we’ll take over the entire market because the other two will churn everybody out and we’ll just sit here quietly and keep our customers by doing right by them and by treating them well and by answering the phone when they call us and fixing all the things that we’ve broken.

Gerry:

That’s a bit radical.

Joel:

Yeah, I don’t understand why nobody has taken that approach. In the airline industry here, an airline called JetBlue pretty much took that approach. You know, let’s not spend money on advertising, let’s just keep our customers by treating them well, and their businesses have grown steadily as a result.

Gerry:

To shift tack again somewhat, you’ve talked about the power of taking the blame. Do you want to tell me the origin of that idea for you, and don’t you think that taking the blame for something that’s genuinely not your fault is in some ways dishonest?

Joel:

Well, if it’s genuinely not your fault, then I guess you know… [Laughs.] It probably is your fault. The origin of that for me, I’ll tell you a little story here. One morning I needed an extra key to my apartment and I stopped by the locksmith who was a couple of blocks away and asked him to make another key and he did something and gave me a key and it was very important to get it right because I was giving it to my mom. So I went back to the apartment, tested it, and lo and behold it did not work. And this happens sometimes so I went back to the locksmiths and said; “The key you gave me didn’t work,” and he said; “Okay,” and he made another one. And at this point I’m late for work and I went home and tried it and the second one didn’t work either, and I went back to him the third time, literally I was on the verge of just giving up on this guy altogether, writing it off as a loss, keys are not very expensive, never doing business with him again, trying to find another locksmith. But somehow I didn’t think that would necessarily solve my problem. For whatever reason, I had the patience to go to him a third time and I was all ready to start bawling him out and I said; “The stupid key you gave me didn’t work,” and I gave it to him and looked at it and he said; “Oh, that’s my fault”. And when he said “that’s my fault” I felt somehow this enormous feeling of all the anger draining away, I was suddenly not mad at him. I’d walked in ready to punch him out and start screaming at him, and those three words sort of magically defused me in a very stunning way. And I realised that… words can have a lot of power in that particular way, something I’d never noticed. And I don’t know if he was doing this because someone had trained him to for customer service reasons, or he had just randomly realised that he made a mistake and that was the first thing he blurted out. And I think in a litigious society, especially in the United States, people are afraid to take the blame for anything because they somehow think this is more likely to get them sued or whatever, that they’re, you know, assuming some kind of liability.

Gerry:

It’s interesting, two interesting parallels with the health care industry… Kim Vicente talks about the difference in the approach in health care and in the airline industry in terms of being able to take the blame and learn from it.

Joel:

The question is will you then find yourself sued if you take the blame. And actually, if you know a little something about human nature and you realise that taking the blame will actually defuse people, I feel like for however many customers you have for every 100 you’ve made angry, if you defuse 98 of them by taking the blame, you’re going to vastly reduce the number of lawsuits that people hit you with. It’s just a better way not to get sued is by actually making people happier and making them not angry with you. And I don’t see how saying it’s my fault is necessarily going to stand up in court as proof that you knew that you did something wrong and therefore you’re liable.

Gerry:

Now surely technical support people can’t always be expected to be polite and friendly, you’re sort of painting this image of a very polite workforce there. I mean, some of the people who call in are just obnoxious.

Joel:

We don’t have that many. I was actually kind of surprised by that. It’s true that people are occasionally obnoxious or that their emails may seem curt. With email obviously it’s a problem of tone not being conveyed well in an email message. Before I wrote this article I went and asked all the people that have talked to customers here at Fog Creek if they had any advice for me and I said, specifically I said; “Do you have any good tips on how to calm down a customer who’s really angry about something?” and they said; “You know, we don’t really get angry customers,” and it occurred to me actually that for the most part if you’re helping people, if you’re actually going to be able to help them and then you’re treating them in a way that they perceive as being just and fair, they won’t be angry, for the most part. They’re only really angry when they’re calling for the 14th time, because they’ve been screwed around and the previous people that they talked to said they would solve their problems and didn’t solve them.

Gerry:

Tell me about the puppet.

Joel:

The puppet came from a conversation we had here… One of the problems with dealing with an angry customer or a customer who is rude, obnoxious or in some way personally insulting or anything like that. I think probably the biggest mistake you can make when you’re doing customer service is to take that personally. Because they’re not really angry with you as a person, they don’t know you as a person, they’re angry with the institution, they’re angry with the company that they’re calling. If you’re Fog Creek software and they’re calling and they’re angry, they’re not pissed off at you know, you, Joe tech support guy, they’re pissed off at Fog Creek for some reason, and if you take it personally and respond personally it’s only going to escalate and it’s not going to make anybody happy and it’s not going to serve the needs of the business. But also you’re going to feel terrible because you got into a big fight with somebody. So one of the things that we suggest doing is imagining, when you’re the customer service person, don’t imagine that the customer is talking to you, imagine that they’re talking to a puppet that you control, you’ve got a little hand puppet, and that’s who they’re talking to, and you as the puppeteer, your sole responsibility is not to take this personally because they’re insulting the silly puppet. As the puppeteer your responsibility is figure out what you want the puppet to say that’s going to make this customer happy again. And if the puppet loses face, that doesn’t mean that you as the puppeteer loses face. It means that you as the puppeteer have actually done exactly the right thing, you have been successful in defusing the customer in some way. So, in a sense, it’s a part of maintaining your own personal sanity by removing yourself from the debate and feeling like you’re not in an argument with a customer right now, the customer is in an argument with this hypothetical creature and your job is just to figure out what words should the creature say, should you make the puppet say that will defuse the customer.

Gerry:

Joel Spolsky thank you so much for joining me today on the User Experience Podcast.

Joel:

Thank you for having me.

Published: March 2007

A note on the transcripts

We make verbatim transcripts of the User Experience podcast. We then edit the transcripts to remove speech-specific elements that interfere with meaning in print (primarily space-fillers such as “you know…”, “um…”).

Gerry GaffneyCustomer support: an interview with Joel Spolsky

Comments 1

  1. Fred Harrington

    “They had no interest whatsoever in improving their process.”

    Oh man… this… this is the most frustrating thing of all time. Whether it’s a business I’m a patron of or a business I work for. It’s so annoying to see a seemingly obvious way to improve something, you tell them about it, and it’s obvious they don’t even care at all. It’s like, why do I even bother? Ugh.

Leave a Reply

Your email address will not be published. Required fields are marked *