In our work to help companies understand the value of applying process to their marketing and sales activities, we have found many reasons for resistance, including the preconception that much of what goes on in marketing is art and therefore resistant to process management.

In every case, we have discovered that, indeed, some marketing activities—in reality, many business activities—require "artists" to produce the desired output. However, that by itself does not make the output art.

Van Gogh, Monet, and Picasso were artists who produced art. That was their output. Some art today is produced more mechanistically and yet, nevertheless, is art, albeit not produced by artists. In virtually all business processes, the desired output is not art, and some of that output is produced by artists.

That distinction is important, as evidenced by a recent Harvard Business Review article that asks the question, "When Should a Process Be Art, Not Science?" (March 2009, pp. 58-65). The authors argue that "there are some processes that naturally resist definition and standardization." They suggest that good process-management principles rely on standardization, yet some processes should not be "standardized."

We applaud efforts to help the reader understand the need for flexible, adaptable processes in many environments. However, to suggest that such processes cannot be defined is shortsighted.

Although we agree that many people, even at the MBA level, teach process standardization, that does not make it right. Rigid, standardized processes are appropriate when necessary, but to suggest that all processes should be rigid and standardized makes for the creation of many bad processes.

Our resistance to the primary focus of that article lies, respectively, in semantics and a premise: the use of the words "art" or "artistic," and the suggestion that sound process-management principles cannot be applied to processes that require flexibility and human input.

The authors cite several examples that, contrary to the authors' view, reinforce that good process management was in fact followed. One example is that sound process management requires feedback mechanisms.

The authors cite luxury-hotel-chain Ritz-Carlton as an example in which a rigid process gave way to an artistic process.

In truth, Ritz-Carlton followed good process management. The staff used feedback to determine that the process being used was not sufficiently adaptable to meet customer requirements. The process was adapted to allow personal judgment to be situationally applied. The result, according to the authors, is happier customers. How they know: feedback.

However, such a flexible process requires additional skills of hotel staff, else the process will fail to "impress the customer."

Could that be defined as art rather than process? We think not. We suggest they have defined a process where particular skills are required. Whether those are skills of an artist is debatable and not relevant.

We agree that all processes that involve human beings require that those human beings have appropriate skills to work successfully within the process. If "art" or judgment is part of the skills required, then appropriate people must be hired.

The authors cite piano "voicers" as an example of art giving way to process. Again, piano voicing is actually an example of a process that requires an artist with very precise skills. If those skills were not learnable, then voicers would no longer exist. If those skills are extremely rare, then the cost of applying those skills to the one-of-a-kind pianos they create simply makes those pianos more expensive. However, piano manufacturer Steinway & Sons understands the process for producing those pianos and uses an appropriate artist within the process.

Though the example may seem "crude," voicers are really no different from having a complex, high-precision piece of machinery on a production line versus a less-precise piece of machinery.

The highly complex machine is used, when necessary, in the process. The voicer is used within the process to create the piano that the customer requires. The fact that the customer's needs are unique and hard to communicate simply make the voicer's job more difficult, and the process to complete the work more flexible and adaptable, especially when customers change their mind.

Returning to the Ritz-Carlton example, we are sure that Ritz-Carlton, using good process-management practice, provides additional training or reassignment to employees whose judgment fails to produce the desired results. To suggest that is an artistic process misleads the reader into believing it is a person-dependent process. It is not.

The process requires hundreds or thousands of people to do it well, and if someone is sick, another person must take that person's place and perform flawlessly. The process depends on people with sufficient skills to execute it well, but that is not the same as being person-dependent. It is simply a process that is highly flexible and adaptable to the environment. Good processes should be as flexible and adaptable as necessary, and no more.

The movement by some within the business-process-management community to insist on "standardized processes" is flawed, driven by a mindless focus on efficiency over effectiveness. Efficiently producing the wrong output is not a good process, no matter how efficient it may be.

The authors suggest process is about removing variation, and they are correct. They then argue that some variation is good because that meets customers' needs. That is also correct. What you have is simply the difference between value-added variation and non-value-added variation, which, again, is part of good process management.

Non-value-added variation should be removed, but to suggest that in all things customers want exactly the same thing is to misunderstand what the customer may value.

To further understand the concept of flexible processes, let's consider the three steps that the authors describe in the article for managing art (as opposed to managing process):

Step 1: Identify what should and shouldn't be art. They suggest that making such a determination is difficult. It is not. Each process is perfectly constructed to produce the results it does. If the process is not sufficiently flexible to produce appropriate results, it is a bad process. That does not suggest art; rather, it suggests appropriate flexibility.

Alternatively, if the process requires artistic skills on the part of one or more of the people involved in the process, then you have to make sure people with those skills work in the process. If the process inhibits the performance of those people, it is a bad process.

Step 2: Develop an infrastructure to support art. That is unnecessary. Again, good processes are sufficiently flexible to produce the required output, and no more. They also include appropriate feedback mechanisms to know if and when the output is no longer meeting the requirements, so the process can be adapted (to wit, what Ritz-Carlton did).

And, as mentioned previously, the process infrastructure must support the people in the process, whether they be artists or not. The authors also suggest that the infrastructure have appropriate metrics. We agree 100%.

However, they suggest that internally focused metrics are appropriate for some processes. That is what did in reengineering. Metrics fall into two categories: control metrics and quality metrics. Control metrics are internally focused to ensure the process is in control. Quality metrics must focus on the ultimate user of the process output.

The authors then suggest that you have to get art and science to play well together. But if you manage processes appropriately, that is not necessary. Good process management is good process management.  Which is where we find our biggest disagreement with the authors, because they have constructed a straw man that exists only because too many processes are badly designed or badly managed.

Step 3: Periodically reevaluate the division between art and science. Again, we agree but insist that good process management is at least about ensuring the process is producing the results intended. Efficiency without effectiveness is waste.

It is not about art, as that suggests process management is not appropriate. It is about flexible, adaptable processes. What they call "art" we call "flexibility and adaptability."

And, of course, some processes require human beings who may be considered artists. That does not suggest they cannot work in a properly defined process. Maybe that's semantics, maybe it's more. However, one thing we know for sure: Managing processes is easier than trying to manage artists.

Sign up for free to read the full article. Enter your email address to keep reading ...


Mitchell Goozé is a principal with Customer Manufacturing Group (, a business-process-improvement consulting firm focused on the demand chain. He is also the author of three books on marketing. Contact him at or at 408-496-4585.