But World Class organizations have moved away from mass production thinking, the principles of systems thinking being far superior in terms of customer responsiveness and operating efficiency (for the best documented manufacturing example see "The Machine That Changed The World"(1) Link: ).

Perhaps the most significant issue for managers is that few are aware of this shift. While the logic of mass production is plausible, the consequence is an organization form that is not especially 'customer friendly' and not especially efficient.

Planning
The starting point for planning a Call Center is working out the volume of work to be done. Consistent with mass production thinking, this is done by counting call volumes and duration. Typically Organization and Methods people take measures of current calls, wherever they occur in the organization. From this, the number of people required to handle the volume of work is calculated.

In simple terms, this tells managers the number of people required in the new 'factory'. Decisions are made about the 'acceptable' level of waiting time, abandon rate and wrap-up time. These decisions then become standards and targets for the number of calls per agent, time to answer, call duration and so on.

Management sees its task to be managing the work according to the plan, the principle tools used for managing being the budget and work standards. Managers become preoccupied with productivity measures, for example cost per call. While the rhetoric of Call Center management is service, the prevailing focus and methods revolve primarily around cost.

Planning
Top-down, resource-based

Design
Functional specialization

Internal Measures
Activity, cost, productivity

Customer measures
Service levels

Management ethic
Manage budgets and people Fig 1. Mass production principles in a Call Center environment (typical in most call centers)
copyright John Seddon

Signs of failure
Paradoxically, the first signs of failure are costs going 'out of control'. For example, one organization we studied planned for three call center factories and got to five before the chief executive started to express alarm. Managers rationalized the reason for the apparent excess volume of calls, arguing the phenomenon to be comparable to the unexpected volumes of traffic on a new beltway, for example; because it was there it attracted more traffic!! But the managers knew nothing about what was causing the increased demand. Their rationalizing was preventing them from learning.

Analysis showed they had caused the problem themselves. The primary cause of unexpectedly high call volumes was a failure of the organization to deal with customer demands, exacerbated by the decision to build call centers.

As customers' demands remained unsatisfied, the customers made repeated calls. As much as fifty percent of the work of the call centers was dealing with such 'failures'.

To understand the causes, managers need a different perspective. They need to learn how to understand their organization as a system.

Systems thinking
The work of a call center should not be thought of as managing units of production (the number of calls to be taken,) but as managing value to the customer (how the organization responds to customer demand.) If a customer experiences exactly what he or she requires and no more, service and efficiency will be optimized. Good service really is cheaper. It is only from this point of view that call center operations can be understood and improved.

To understand the work this way requires management to start by looking at the organization from the outside-in. This means understanding the nature of demand; in particular, assessing what is predictable about why customers call.

The Call Center as a system
While Call Centers have been designed on mass production lines, they are in fact systems. Call centers do not 'do as they are expected' because they 'do as best they can' in the circumstances. The nature of customer demand upon a call center is predictable and needs to be understood. Much of the demand is caused by other parts of the organization, for example.

When managers learn just how their call center interacts with the rest of the organization - just how and how well the parts work together - they find for themselves the means to improve performance. Action on the system leads to reductions in cost and improvements in service, neither of which are achievable by focusing on standards and costs - the tools of mass production management.

A distinction we use is 'failure' demand versus 'value' demand. Failure demand is demand caused by a failure of the system, for example 'it did not arrive', 'it's not what I ordered', 'you didn't get back to me' and so on. Value demand is what the organization is there to do, for example 'give me service', 'can I buy', 'which is better' and so on.

In our research we have found as much as seventy percent of demand on call centers to be what we describe as 'failure demand'. To date, the least 'failure demand' in any one case has been twenty five percent. This figure suggests that the productivity of all call centers can be improved by at least twenty five percent.

Failure demand
Customer progress chasing is an easily understood example of 'failure demand', but there are many others.

To understand the causes of 'failure demand', managers need to understand the interdependencies within their organization. This is often difficult for managers to do. The call center having been established as a function in its own right, managers experience the same constraints about working across boundaries as do managers in any other part of the organization.

Boundaries, in effect, protect 'other people's patches'. Indeed, in many cases we have found organizations creating cultural walls between their call center and the rest of the organization, as the call center was established to take on work that used to be done elsewhere. As the planners see it, to open such boundaries to problem solving might result in a return to the very position they sought to move away from.

However, the primary block to learning about the interdependencies that cause much of the demand on call centers is the assumption that all demand should be treated as units of work. There is no distinction with respect to type of demand and thus no facility for understanding how much of the work of a call center is being caused by the organization itself. Which, in turn, means that the cost of servicing 'failure demand' is never an overt factor in management's equation.

The next step in optimizing a Call Center's operations is to understand 'value' ? what matters to the customer in every type of demand placed on the system - and 'flow' ? how the organization responds to the demand. Failure to understand 'value' from a customer's point of view and failure to improve the operation of the organization's activities (flow) for the fulfillment of that value perpetuates sub-optimization or waste for both customer and organization.

Demand ?value ?flow

As managers discover when they study value and flow, what happens in practice is quite different from what was planned and what was planned is often not what is required. The people who do the work know this only too well. Rather than find out what is going on in ways that will lead to improved flow, which means reduced costs and better customer service, managers are prone to behave in ways that exacerbate the problem. We have found a variety of types of waste built into call centers by managers attempting to manage the work as 'managing units of production'.

Some examples:

progress chasing calls (and other types of 'failure demand') being treated as 'normal' demand ? units of work - thus consuming resources

functional specialization leading to calls being passed to different levels or groups

fragmentation of value: knowledge of the customers' problems being lost or distorted as the call moves between agents and/or levels

setting work standards

using incentives

prescribing procedures for dealing with customers

Act on the system (not the people)
The mass production manager sees his or her job as setting and monitoring work standards and procedures. 'Doing as you are told' in the face of variation in demand is a primary cause of poor morale. Are all customer calls the same? Are all products and services the same? Does dealing with 'failure' take the same time as dealing with 'value' work? These are all causes of variation in the call center service agents' lives.

It is another of the uncritically inherited assumptions of mass production thinking that people are the primary cause of poor performance rather than the system in which they work. As a consequence, management becomes concerned with people meeting work standards or targets, which are derived from the plan and therefore must be achieved.

The ability of operators to achieve targets or standards is governed by the system in which they work. If, as is so often the case, people cannot meet the target or standard they are left with no option other than to 'cheat' in order to avoid grief, and the methods people employ to avoid grief is a testament to their ingenuity. A good system would harness ingenuity towards purpose rather than against.

All performance is subject to variation. The question is - how to find the point of greatest leverage for sustained improvement? To accept that the system has more influence on people's performance than differences between the people themselves, managers first need to learn a little about the theory of variation. Needless to say, variation, as distinct from variance, is not part of the mass production legacy.

Finally, the failure to appreciate demand in customer terms leads to a failure of 'preventative' training. People employed in call centers are trained in 'internal' procedures - 'what procedure to follow when you get one of these'. A thorough understanding, from the customers' point of view, of why customers are calling in always leads to different training, focused on creating value for customers.

In transforming a call center, management's focus changes from managing people (ensuring people are doing as they should) to managing the system; understanding and improving how well the work flows, end to end, to fulfil the customers' expressed value.

Achieving transformation
The prerequisite to successfully transforming a call center's operations is to learn how the current organization works as a system.

Perspective
Outside-in

Re-design
Demand, value and flow

Internal Measures
Variation and capability

Customer measures
What matters to customers

Management ethic
Act on the system Fig 3. Principles for understanding and managing a call center as a system
Copyright John Seddon

The starting place is to understand the type and frequency of customer demand. The only reliable method we have found is sitting with people who work the 'phones, working through a series of iterations, listening and classifying. At the start of such exercises managers often claim their calls will be entirely unpredictable. On the contrary, we have found every call center's work to be largely predictable.

Often, managers will ask us to use the information they have from their call coding, their staff being told to follow procedures for coding calls. We have always found such information to be unreliable, since agents are more concerned to meet their work standards than accurately code their calls, plus call classifications tend to be derived from 'internal' classifications rather than customer classifications. The next requirement is an understanding of the extent and causes of variation in current performance. This exercise always reveals the impact of the system to be far in excess of that of the operators.

When managers discover the opportunities to improve flow by action on the system, they recognize for themselves the fallacy of managing through budgets, standards and targets and learn how important it is to act instead on the causes of variation. Clearly there will always be differences in performance attributable to operators, but in comparison to the variation caused by the system, these are minimal.

In every case, when this program of transformation has been followed, action on the system has produced outstanding results. In a computer service organization (a relatively complex call center), the results were reductions in time to resolve customer problems, reductions in waste and improved morale. In a cellular telephone service provider (a relatively simple call center), the results were a drop in abandon rates from over sixty percent to zero, a drop in the number of calls coming in from customers and, again, improved morale.