In the last few years a lot has been written about Business Process Management, and about technologies supporting it such as BPMS, SOAP and Web Services. Most of these theories, tools and techniques refer to processes of a highly structured nature.

Typically, BPM theorists and practitioners have focused on highly structured processes, like back-office processes of industrial or administrative nature. These processes are highly standardized and repeatable, produce a consistent output and are likely to be automated in part or end-to-end (STP). All process instances are executed in a very similar way and it is easy to draw a flowchart detailing the sequence in which tasks are executed. It is also possible to formalize the business rules that guide decisions, normally based on the evaluation of some process variables.

But recently other kinds of processes have caught the attention of process management specialists. They are known as knowledge processes, or knowledge-based processes. Knowledge processes can be defined as "high added value processes in which the achievement of goals is highly dependent on the skills, knowledge and experience of the people carrying them out". Some examples could be management, R&D, or new product development processes.

Knowledge workers carry out these processes by taking into account multiple inputs (generally a wide set of unstructured data and information) to perform difficult tasks and make complex decisions among multiple possible ways of doing the work, each one implying different levels of risk and possible benefits. They are dependent on individuals and it is not possible to automate them.

One example of a knowledge process is "Marketing a new product". The same steps are followed each time a new product is launched (benchmarking competitors, deciding pricing strategy, planning promotion, etc...), but it is the experience, knowledge and intuition of the people that drive the process to success.

* Multiple inputs to the process exist

Some of them would be competition, lifecycle stage of the market, brand image, budget, etc...

* Complex decisions are made

There are many possible ways to achieve the process objectives (reach planned sales, leverage brand image, etc...)

* Each decision implies different levels of risk and potential benefits

It is the responsibility of the worker to choose the best one (low price strategy, aggressive advertising campaign, etc...)

There are three main characteristics that make knowledge processes different from highly structured processes:

Focus is on communication instead of automation

The key to process improvement is to clearly communicate process definitions (the way in which the company wants the processes to be carried out) to the people in charge of their execution (through training, process descriptions publication, etc...). The better process participants understand the process definition, the higher the probability that the process is carried out according to it.

They are better implemented through obtaining buy-in than through imposing directives

They are more difficult to implement through discipline than administrative human-centric processes (although some discipline is needed). It is better to focus on obtaining buy-in from the people affected by the processes through early involvement, communication and expectations management. It is a known fact that knowledge workers are reluctant to change their habits. Some say knowledge workers don't like following procedures because they feel it limits their creativity; but most of the time they will be happy to follow a procedure as long as they see value in it, perceiving that it helps them work better and produce a better process output.

Process definitions are high level descriptions instead of rigid workflows

Processes can only be defined up to a certain level of detail, and it is difficult to provide low level work instructions or to automate decisions. Because they cannot be formalised in detail, process simulation is rarely possible. Decisions are highly subjective and too complex to be expressed in a formal language, as they are taken based on intuition and not on rigid business rules.

It is extremely important to continuously improve knowledge processes, by creating an environment through which they can evolve. This can only be achieved through coordination of diverse disciplines such as knowledge management, change management, expectations management, etc... It is crucial to establish an adequate process context (the combination of technologies, procedures, people, etc... that support the processes). The process context must incorporate feedback mechanisms, change evaluation procedures, process improvement methods and techniques and must be flexible, in order to be able to incorporate enhancements in an agile but controlled way.

If the process is instantiated frequently and the instances are homegeneous, it is possible to create great process models that dramatically increase the efficiency of the process. The best way to ensure process improvement is to generate an environment in which people are motivated, enthusiastic and passionate about process management.

Most of the time, knowledge processes are collaborative. By performing a process collaboratively it is possible that each task is carried out by the most specialised, experienced and knowledgeable worker in that specific area. Having a net of relations within the organization is a very important asset for people executing knowledge processes.

In the last years some organizations have emerged with the aim of creating professional communities around specific disciplines such as Software Development (SEI, ESI, etc...), Project Management (PMI), Business Process Management (BPMI), IT Service Management (ITSMF), etc... One of the objectives of these groups is to develop a body of knowledge that compiles the discipline's best practices in the form of reference frameworks, methodologies and maturity models. These assets should be considered by any organization interested in knowledge process management.

It is usual that knowledge processes take the form of projects to manage their execution. If the output of the process is a unique product, managing work as a project will result in obvious advantages.

There are certain guidelines that can help an organization willing to improve their knowledge processes:

* Provide process description on how to approach work

Try to figure out the best way to carry out a knowledge process, by making the best practices existing in your organization (or in your industry) explicit. Publish process definitions in a format that is easy to consult and understand.

* Provide tools that facilitate and standardize work

Decide which tools are best to help knowledge workers carry out their work. Involving all affected knowledge workers in the process of deciding which tools will be used is very convenient, in order to obtain user buy in. It is a good idea to choose a champion for each tool who will master its use.

* Assign owners to processes

Choose a person with leadership skills and the appropriate level of responsibility and influence and make him/her accountable for continuous improvement of the process. Give him/her a clear objective to achieve and an incentive to reach the goal.

* Encourage feedback for process improvement

To ensure that the flow of information between executors and the process owner is fluid, encourage people to contribute to process enhancement through incentives. Use your imagination to reward contributors (consider not only monetary incentives).

Lucas Rodr?guez Cervera is founder of Nevant ? Process documentation software, a company specialized in delivering process solutions to knowledge based companies. They pioneered this concept with metoCube.