How to Design Poor Service – Expect 100% Utilization of People or Resources
I have as many bad customer service experiences as the next guy, both in healthcare and other businesses. As I've said before, I try not to be a “hack” blogger who just uses their platform to complain about the last bad thing to happen, unless there's a broader lesson involved.
Let's say you HATE your customers. Sounds unimaginable, but I've long suspected this is actually the case with American Airlines, having flown almost a million miles with them. OK, so that's a bit harsh, but it does seem true that American falls into the trap that many other companies do – expecting 100% utilization (of equipment or people) in the name of efficiency, while ignoring the laws of nature that say 100% utilization leads to LONG customer waiting times – in person, or on the phone. I've also had similar experiences with Verizon FIOS… bear with me, I'm not just griping and moaning in this post.
If you've studied queuing theory, Industrial Engineering, or the famous textbook Factory Physics, you know that 100% utilization of a resource is a recipe for long waiting times and poor customer service. It's a mathematically provable law of nature. Expecting to have 100% utilization of a machine in a factory, means you'll have lots of waiting time and lots of inventory (and/or poor customer response). Expecting 100% utilization in a call center means lots of busy signals or customer hold time.
Long story short, American Airlines misplaced our suitcases for the first four days of our vacation back in June. That's hardly breaking news that airlines lose bags. My wife and I were more upset about:
- The finger pointing that took place between American and Iberia (our transfer was in Madrid)
- Failure to deliver on the promised response (they didn't bring our bags to our hotel, as they had promised)
- The insincere “we care about you” mass emailings that American calls a “personal communication”
So when we got back from vacation, we wrote a letter to American, complaining about the generally indifferent and surly service that we encountered at every step of the journey (again, this isn't a news flash, this is like a hack comedian complaining about bad airline food).
The First Instance of 100% Utilization – American
Within about a week, I received a phone call from someone in the Executive Office of American Airlines. OK, that's not bad. But I was on the phone for work, so she left a voice mail with her name and number.
I called back later that day… no response.
I've called back four or five times over the last two weeks and I have always gotten her voice mail with the message of “I am already on the phone with another caller, please leave a message.”
Granted, it's a limited sample, but is she busy making outbound apology calls all day long? Without being too obsessive and dialing non-stop or something, I find it hard to believe that somebody is so busy that they can't return a call — they only have time for one attempt?
No amount of “we care” emails or a “we received your letter and we care” postcard can gloss over the fact that American apparently doesn't have enough capacity to deal with the number of complaints they receive. Instead of just adding more reps, ultimately, their goal should be to reduce customer complaints so fewer reps are needed. But that's really pie-in-the-sky thinking for American and the other old legacy carriers.
I don't know if I'll ever get through to that Executive Office person. I'll let you know if I do. I'll try again Monday morning.
The Siren Song of 100% Utilization
Industrial Engineers learn that 100% utilization is bad. According to the lessons of Factory Physics, depending on the system – how much variation you have – waiting times start to skyrocket after 80% utilization or so. If your primary objective is low cost, you'll drive 100% utilization, and customers will suffer. If you really want to be responsive, you can't plan for 100% utilization. Looking at somebody being “not busy” is a very obvious cost – and hating to see people idle is old Frederick Taylor thinking.
Lean thinking, going back to Taiichi Ohno, would hate to see the part in the factory not moving. That's a different type of waste and a different focus. In a service setting, seeing the customer wait would be worse than seeing an agent idle, if you're really trying to be Lean. It's easy to calculate the cost of an idle call center agent. It's much more difficult to determine the cost of an angry customer caused by poor customer service.
It seems that general MBA thinking teaches managers to focus solely on cost, when the real problem (not just in my selfish case) is poor flow. American Airlines, like other airlines, is obsessed with 100% utilization of the planes, which works fine unless a flight gets canceled… you need slack capacity in any system if you're going to have flow.
Another Case of 100% Utilization – Verizon FIOS
A few weeks back, I switched from my local cable company to Verizon FIOS for phone and internet. I could get a faster connection for a lower price, so it seemed like a great value.
The wireless router that's included in the hardware was almost constantly fritzing out, requiring reboots. After a number of customer service calls over two weeks, I finally convinced a phone agent to send me a replacement box.
The decision to actually send the box required a 2nd-level tech support approval. The 1st-level agent I was on the phone with apologized for the delay waiting for the 2nd-level tech to be ready to talk with her via Instant Messenger.
After waiting (I'm sure the 2nd-level tech was being kept 100% busy, being a more expensive resource and all, typical MBA thinking would say), the 1st-level tech (1LT) and I and the 2nd-level tech (2LT) had a discussion like this:
- 1LT: What operating system are you using?
- Me: Well, one machine is Windows 7 Professional, the other is Mac OSX Snow Leopard, why does that matter?
- 1LT: (typing furiously)
- Dead air — waiting
- 1LT: Have you tried changing the wireless channel?
- Me: Yes, many times, that's well documented from the previous calls
- 1LT: (typing furiously)
- Dead air — waiting
This continued for a few more questions, with very long waits before the next question. I asked the 1LT what was going on.
She said the 2LT was always on IM with four different 1LTs and four customer issue at the same time, so due to the multitasking we had to wait.
It seems the ONLY reason you'd design a system like that is if your only concern is 100% utilization. That's cost focus (it makes me want to pull my MBA diploma off the wall). My call took at least four times as long as it should have and it was ten times as frustrating. Keeping the 2LT busy wasted my time and the time of the 1LT.
Can you imagine a system design like this in a physical setting? Does a grocery store drive efficiency by having one cashier run back and forth across four registers? Does a hospital try to have one rep register four patients all at once, running back and forth across different desks.
Why does this job and system design seem to make sense in a virtual setting? Are people truly allowed to have joy and pride in their work, as Dr. Deming would say, when they're kept 100% busy like this?
The only conclusion you can draw is that these companies are more concerned about silo-ed cost than they are about keeping customers happy and having good flow. Thank goodness NOBODY is holding up either American Airlines or Verizon as a company that's supposedly using Lean principles. There's nothing Lean about a system that mandates 100% utilization rates.
Again, it's far more concrete to measure the cost of an idle person than it is to measure the cost of poor flow, unfortunately. That drives a lot of bad short-term decisions, apparently.
A really good read on the topic of slack, utilization, and flow is Slack: Getting Past Burnout, Busywork, and the Myth of Total Efficiency.