Service companies can’t show customers a tangible product. Since services are intangible, the only way to sell them is by making a promise to perform. But most service companies fail to keep their promises, leaving customers frustrated, confused and abused. Why do so many service companies fail to keep their promises to customers?

Customers have become accustomed to being abused by the companies they buy services from. Customer expectations for most services are low, low, low.

Fewer than 1 in 10 customers say companies are exceeding their service expectations. Customers are not getting the service they want. Many feel abused. When they call your customer service line and hear a recorded voice saying “your call is important to us,” guess what? They don’t believe you.

The most hated companies, and the most hated industries, are service providers: Consider your last experience with your telephone, cable or satellite provider; your utility provider; your airline; or your bank. Did you talk to a real person or an automated system? Did you have to ask to speak with a supervisor? Did you lose your temper? Did you swear? Hang up the phone in disgust? Did you not even call at all, because you knew it would not be worth the time and effort? You are not alone.

The customer support decision tree.

When customers are dissatisfied, there are only two possible outcomes. Either the issue is resolved to the customer’s satisfaction or it isn’t. Trying to reduce or outsource your customer support costs might actually cost you a lot more in the long run, as you lose customers and they badmouth you to all their friends.

Consider what happens when you focus on costs in your customer service operations: The cheapest customer support call is the one that doesn’t happen at all. Of course, you don’t know why the customer didn’t call you. Maybe they were happy with your service, or maybe they just didn’t have the time or energy to fight their way upstream through your draconian support system. There’s just no way to know.

If you’re lucky and they do call you, there are still two possible outcomes. You satisfy the customer or you don’t. Probably the cheaper outcome from a functional, departmental, cost-accounting perspective is the one where you quickly tell them it’s not possible or it’s not your department. You can mark that “resolved, not my department” and forget about it. Maybe the customer will give up at that point.

If you’re still lucky, your customer will be patient as you bounce them around your system till they find the department that might be able to solve their problem. Once again you have two possible outcomes. You satisfy the customer or you don’t. Of course while the customer bounces around they might get a little frustrated, but that’s okay, because that’s not your cost, it’s the customer’s cost. It’s an “externality” and it doesn’t show up in your cost-accounting system.

Now that the customer has finally reached the right department, you still have two possible outcomes: you can resolve their problem or not. It’s probably going to be cheaper not to solve their problem, because most of the time solutions have some kind of cost. You might have to accept a return or credit the customer’s account. If the situation is not in the service rep’s rule book, the service rep might not be able to help.

If you’re lucky, the customer will ask to speak to a supervisor. At that point there are still two possible outcomes. You can solve the customer’s problem or say no. It’s probably cheaper to say no, for the reasons outlined above. But maybe the supervisor will say yes, in which case there’s a chance your customer will be satisfied.

Customers will put up with this kind of treatment only for so long. Eventually they will find another company that treats them better. That would be great for you though, right? Because the cost of serving someone else’s customer is zero! Yay for the cost savings team!

Then, when your customer tells all their friends about their experience, you will lose more customers. Your call center costs will continue to go down. At some point when the last customer has left, you can eliminate your call center altogether. Total cost victory achieved.

Why do companies have such a difficult time providing good service?

Many service companies just aren’t designed for service delivery. They are designed like factories, optimized for the mass-production of inputs into outputs. This makes perfect sense in a rapidly-industrializing economy. But in an economy where manufacturing is shrinking and services are expanding, it doesn’t work anymore.

Traditional management thinking looks at a customer service call as an input to the “service factory.” Most companies try to standardize these inputs as much as possible so they can process them efficiently. The factory’s job is to produce “resolutions.” This is how we end up with complicated voice menu systems that attempt to route calls to the appropriate department while keeping costs as low as possible.

For a factory, it’s not difficult to get standard inputs from suppliers. But inputs from customers come in all kinds of different shapes and sizes. Every problem, every job that customers need to do, has its own unique profile.

As companies try to fit customer demands into standard boxes, customers become frustrated and angry. They give up. Sometimes they leave to find another provider, but even then they often hold little hope that anything will change.

What’s wrong with the way companies are organized today?

In a service-driven marketplace, the focus needs to shift from the line of production to a different line; the front line. The line of production is a one-way arrow, starting with raw materials and suppliers and ending with the customer who buys the product.

But the front line is not a one-way arrow, so much as a boundary, like the cold front in a weather pattern: it’s the edge of the organization, the interface where customers and the company interact.

Optimizing for the production line and optimizing for the front line require fundamentally different kinds of organization.

A production line requires efficiency. Inputs can be standardized, and environments and processes can be internally controlled.

But a front line requires optionality. Front line people deal with environments and circumstances that cannot be predicted. What they need are support systems that they can access as needed, like “call a friend” on Who Wants to be a Millionaire.

A company that’s organized for production can measure quality internally and objectively. Product quality can be defined by the degree to which a physical object conforms to, or varies from, a standard template.

But a company organized to provide services has no physical product to measure. Services are experiences, and the only quality measure that matters is subjective: how the customer perceives the experience.

Thus, most service companies are focusing on the wrong things: they are doing the wrong things, and they are measuring the wrong things. No wonder customers are frustrated.

Measuring service quality.

Only one person can judge service quality: the customer. Customers come to you with a purpose: they have a job they need to do, and they want you to help them do it.

Service quality is the difference between that purpose – what customers expect – and your performance – what they get. If customers get exactly what they expect, then the service is successful. If you exceed their expectations, customers may be delighted. If they get less than they expect, your service falls short.

Therefore there are two ways to improve service quality: improve your performance on the things customers care about, or get your customers to reduce their expectations to better match what you can reliably deliver.

Marketing’s job is to make realistic promises that help set customer expectations. It’s the job of the rest of the organization to deliver on that promise. It’s all about the match. A company that promises more than it can deliver will disappoint customers, and that’s very dangerous, because today’s customers have the tools to tell the world about their poor service experiences.

A good service is one that can perfectly match customer purposes with the company’s promises and performance. For example, if you go to Amazon, search for a book, click buy, and the book is delivered as promised, then the service succeeded.

Balancing promise, purpose, and performance.

Here’s how it’s supposed to work: Your promise sets the expectation. When a customer has a purpose that matches your promise, they come to you with that job to do. If you can perform in such a way that you meet the customer’s expectations, you win. To succeed in services, companies must align promise, purpose and performance as closely as possible.

Consider Southwest Airlines. The company’s promise is to deliver cheap flights within the US, and to make that no-frills flying experience as fun as possible. Customers come to Southwest with the purpose of getting from here to there as cheaply as possible. And Southwest delivers on its promises. Think of a service provider you’re happy with (I know, it might take a while) and you will find a good match between that company’s promise, your purpose, and their performance.

How do they do it? For Southwest to achieve this alignment requires that they coordinate a whole host of activities to support that simple promise. To take just one example, for Southwest to make the flying experience fun, they must have cheerful workers. To have cheerful workers, they must have great relationships with employees.

Southwest is 85% unionized but leaders at the highest level are committed to good employee relationships, and consequently the company has very little labor strife. Employees, spouses and dependent children fly free on Southwest and enjoy discounts on other airlines. Employees can also earn up to four “buddy passes” per quarter so friends can fly free too. Employees share in the company’s profits (yes, an airline with profits! Southwest has been profitable 38 years running), and enjoy excellent health benefits including vision and dental.

The Law of Requisite Variety.

Since customer needs come in all shapes and sizes, variety is a fact of life in any service business.

The Law of Requisite Variety, also known as Ashby’s Law, states that any control system must be capable of variety that’s greater than or equal to the variety in the system to be controlled.

In other words, if there is variety in the environment you need enough variety in your system to absorb it effectively. Think of a juggler: no matter how skilled the juggler, there will always be a point where there are too many possible states for the juggler’s mind and hands to maintain control.

There are two ways to deal with variety. You can reduce variety by standardizing inputs and controlling the environment as much as possible, or you can design a system that’s capable of absorbing more variety.

Reducing variety.

Reducing variety is a factory approach to service design. This works when you can successfully constrain your inputs to a small number of possible states and control the environment where the service is delivered.

McDonald’s is an example of a service company that succeeds because it is able to reduce variety. Consider the McDonald’s drive-through experience: As a customer your orders are limited by what you can see on the menu. You order some quantity of menu items, which triggers a factory-like production process behind the scenes, while you pay at a payment window. You pick up your food at a delivery window. Purpose matches performance. Service delivered.

As a customer you are a willing participant in the factory experience. Your car is an item in an assembly line and you can be sure that the time between your order and the delivery of your food is carefully monitored by the company.

This works for McDonald’s because they have organized their promise around something that can be consistently and reliably be delivered in a factory-like way. For this to work, customers must follow the process as prescribed. If you walk in to a McDonald’s restaurant, you don’t expect the kind of service you would get in a sit-down restaurant. If you sat down at a table and waited to be served, you would be waiting a long time. McDonald’s works more like a giant vending machine. You wait in line until you get to the machine interface, you put in your order and your money, and the food comes out the slot. This works for McDonald’s because they have narrowed customer expectations to match a factory-like service.

McDonald’s is an excellent service provider in their niche, not so much because they are excellent at service delivery; but rather because they have reduced their promise to a very narrow window, reducing variety in their inputs and controlling the environment as much as possible. There is a cost to this approach, however. People in such tightly-controlled systems don’t have a lot of autonomy and they have few opportunities to exercise creativity at work. So when a McDonald’s employee says “Have a nice day,” you might feel like it came from a robot. Because when people are treated like robots, they act like robots.

Absorbing variety.

Most services cannot reduce the variety of their inputs and control their environment as easily as fast-food restaurants. Customer demand is not usually so easily standardized and regulated. Customers have many problems to solve in their lives. They have many jobs that need doing, and only a few of them can be easily reduced to a small set of standard inputs.

Customers have a tendency to resist standardization. The more you try to standardize their service requests, the more you will piss them off. Not a good recipe for customer satisfaction or long-term business growth.

In most cases, service providers must reorganize to absorb variety rather than reduce or contain it.

Online shoe store Zappos’ call centers are designed to absorb variety. Most call centers look at customer support as a cost. After all, if you have already been paid for a product and delivered it, then you already have your money and any additional effort on your part will cost you money, right? Zappos looks at the equation differently.

Zappos knows that a customer call probably represents a very tiny fraction of their total interactions with the company. Unlike most online retailers, Zappos encourages person-to-person contact. Zappos publishes their 1-800 number on every page of their site. Online stores don’t get a lot of chances for real human contact with customers, and Zappos does everything they can to turn customer calls into positive, human experiences that customers will remember. Their number one goal is to deliver experiences that are so great they are worth talking about.

The first step in creating a great customer experience is hiring the right people in the first place. After four weeks of training, Zappos call center reps are offered $3,000 to quit immediately. Remember, this is for an $11-an-hour job. By offering people money to quit, Zappos ensures that the people they hire are really excited about working there.

At Zappos, there are no “customer service scripts” or pre-set time limits for customer support calls. Reps are encouraged to take as much time as necessary to solve the customer’s problem, and their mission is to provide the best customer service possible. Zappos has a 100% satisfaction guaranteed return policy. After the call, service reps follow up and keep their promises, and send a personal note as part of their follow-up.

Good customers are profitable customers. Zappos treats frequent customers well, with surprises like upgrades from standard ground shipping to next-day air. Making customers happy, says CEO Tony Hsieh, leads to cost savings elsewhere, like marketing. “We let our customers do our marketing” he says.

By making sure they get the right people and giving them the autonomy and authority they need to serve customers, Zappos call centers are designed to absorb variety, not contain it.

Front stage and back stage.

The decision to contain or absorb variety is not simply an either-or choice. Most strategies will require companies to do both: to reduce variety in some areas while absorbing it in others.

When determining how to make these strategic tradeoffs, look at your company’s business ecosystem and ask, “Where in the system do we expect a lot of variety, and where do we expect things to remain relatively stable?” In areas where you expect a lot of variety, say, due to changing customer preferences or rapid technological change, efforts to reduce variety are likely to be wasted: By the time you have perfected an approach, the environment around you will have changed. It’s like trying to build a perfect sandcastle when the tide’s coming in. You’re going to have to rebuild it anyway.

Service designers divide organizations into front stage and back stage operations. The front stage is the front line, everywhere the company interacts with customers. Front-stage operations are often characterized by high levels of variation, because the jobs that customers need to do rarely come in standard shapes and sizes. The back stage is the set of operational services and support that make the front stage activities possible. On the back stage, it’s more likely that companies reduce variety, by standardizing inputs and controlling the environment.

Balancing the front-stage and the back-stage.

In a 2009 paper, Robert Glushko and Lindsay Tabas use the hotel check-in as an example of how the front and back stages need each other.

Imagine you arrive at a hotel and the person at the hotel desk is not able to find your reservation. That is, your reservation was made but somehow lost by the hotel. This is a backstage failure that affects front stage performance. If your reservation is lost and no rooms are available, the customer experience is likely to fail. There is not much the desk person can do other than try to find you a room at another hotel.

For example, consider the way that Amazon balances front stage and back stage operations. Amazon’s front stage is its customer-facing website. Amazon can expect that the level of change on the web is likely to remain volatile for some time. Constant innovation in online services will cause customer expectations to evolve accordingly. So it makes sense for Amazon’s web-development approach to be highly adaptive and flexible, with lots of room for creative experiments and innovation.

But radical, disruptive innovations on the fulfillment side of Amazon’s business are less likely. It’s reasonable to predict that customers will continue to want fast, efficient delivery; and that warehousing, shipping and logistics, because they involve large investments and existing physical infrastructure (ships, trucks, planes, railroads and so on), won’t change anywhere near as rapidly as online services. So it makes sense for Amazon to focus on reducing variety through standards and controls in its back-stage operations, while maintaining maximum adaptability on its front stage with customers. And indeed, Amazon web developers have a very different work experience than workers in an Amazon distribution center, although the company’s cost-focused, thrifty culture is in evidence throughout.

When Amazon bought Zappos it seemed like a strange marriage. At first glance the companies couldn’t be more different when it comes to their approach to customer service. Amazon’s approach is “if the customer is contacting us there’s something wrong. Zappos is delighted when customers call.

Amazon Founder and CEO Jeff Bezos explains the difference:

“Every time a customer contacts us, we see it as a defect. I’ve been saying for many, many years, people should talk to their friends, not their merchants. And so we use all of our customer service information to find the root cause of any customer contact. What went wrong? Why did that person have to call? Why aren’t they spending that time talking to their family instead of talking to us? How do we fix it? Zappos takes a completely different approach. You call them and ask them for a pizza, and they’ll get out the Yellow Pages for you.”

So if Amazon doesn’t want to absorb Zappos’ culture, why buy them? If you look at this through a front-stage backstage lens it becomes clear. Zappos backstage is the same as Amazon’s – customers care about fast and efficient delivery and returns. So it makes sense for Amazon and Zappos to consolidate their back stage operations to achieve economies of scale and operational efficiency.

On the front stage, Zappos gives Amazon greater variety and optionality. Who’s to say if there’s a better business in high-touch or high-tech approaches to serving customers? Well, if you don’t know, why not try both? In this way Amazon has a chance to be both the low-cost Wal-Mart and the high-touch Nordstrom of the web, with the added advantage of a combined, world-class, backstage logistics and delivery system that improves the cost-effectiveness and global reach of both companies.

Service delivery depends on the ability of the front stage to react creatively and proactively to situations as they come up. But it also depends on the ability of the back stage to anticipate what kinds of things will be necessary to support the front stage.

Low expectations mean big opportunities.

In the early 1900s, Henry Ford took a leap of faith that if he could build a cheap, reliable car, people would buy it. His leap turned out to be justified and with the Model T, Ford opened the door to an era of unprecedented material abundance. With that leap of faith, Ford discovered a vast reservoir of latent demand that had previously been invisible.

What is the huge, invisible, untapped market today? Where is the latent demand that will drive growth in the 21st century? Look at all those customers who are unhappy, underserved, and abused. Look at the extremely low expectations they have of their service providers. Tell me there isn’t demand for companies who truly serve their customers, who value their customers’ time, and who are built to deliver exceptional service experiences.

More than half of customers are willing to spend more for excellent service.

Customers have jobs they are trying to do all the time. Jobs come up every day and in all kinds of ways. And most, if not all of them, can be done a hell of a lot better. Today is a good day to start the process of redesigning your company for service delivery. Then you can get into the real business of services: making promises you can keep, and keeping the promises you make.

Further reading.

Designing Services that Deliver by G. Lynn Shostack, Harvard Business Review (January 1984).

Designing Service Systems by Bridging the “Front Stage” and “Back Stage” by Robert Glushko and Lindsay Tabas (April 2008).

The Service Encounter: Diagnosing Favorable and Unfavorable Incidents by Mary Jo Bitner, Bernard Booms, and Mary Stanfield Tetreault, Journal of Marketing (January 1990).

The Profitable Art of Service Recovery by Christopher Hart, James Heskett and W. Earl Sasser Jr., Harvard Business Review (July 1990).

Many thanks to Tom Graves for his insights and help developing these ideas.