On Capgemini's CTO Blog, I posted this blog-item on 'No Process', as part of Capgemini's TechnoVision 2014.
Process On The Fly #3 - No Process
Building on the next generation of Business Process Management, Business Rules, Event Processing and Case Management platforms, new flavors of process can be modeled, executed, monitored and managed. Guided by context-sensitive and analytics-driven support many fixed, inflexible processes can be replaced by concurrently executed activities that optimize the time of human resources and their knowledge by having them ‘swarm’ around cases and results that need to be produced, never following a predefined path. So in the end, the ultimate Process might be No Process at all.For quite some time, process models, procedures and process-based work instructions have been used to attempt to influence the behavior of employees, in the hope that this creates efficient, customer-friendly and compliant processes. And, as part of IT innovations, Business Process Management technology has been used to create process-aware systems, in the form of workflow systems, that coordinate the execution of processes using these process models.
Although these process models have their power, they have two serious problems.
First problem: As an instrument for influencing behavior, process models are not very motivating. Quite the opposite. Most people prefer to leave process models in dusty binders as long as possible. And if they use them, often hesitantly to try to find how to deal with a situation, they frequently get lost (as these models are often incomplete, sometimes quite abstract, and not organized towards the contextual situation-specific approach).
Second problem: Process models are rigid. When a customer request is delivered following the process model, the model is usually deaf and blind for unexpected events, new insights and needs for other activities or paths. Thankfully, most of the time, employees are smart enough to work around the process when needed, but unfortunately we all encounter enough employees who are not.
You might compare the use of these process models to driving towards a certain location using a prescribed route that you brought along on paper. You enter a large traffic jam, but don’t know an alternative. You pass a beautiful forest, but don’t want to leave the route as you are scared to get lost. You realize you want to buy some food on the way, but again are afraid to detour. The process model limits your freedom. And then the inevitable happens: An exit has been blocked due to roadwork. The paper process model is suddenly useless.
Of course, one can attempt to create the all-encompassing model, with all possible events, rules, activities and routes. But this path won’t work. First, the process model will explode in complexity. Second, some situations simply cannot be fully prepared at ‘design time.’ Think of a complex diagnosis and treatment in a hospital – the possible symptoms, causes, possible treatment interventions and patient reaction to treatment. You just won’t know upfront; the process path will emerge, patient by patient.
In modern car driving, the technology (not perfect, but quickly maturing) to better deal with these ‘process model aspects’ is of course the navigation system. Modern navigation systems offer the strength of a process model (direction: what’s the next move to get from A to B), without the limitations:
- Context driven: The system knows where you are, and gives you only the relevant information.
- Responsive: The system offer insights (traffic jams, road blocks) and suggest other routes.
- Goal driven but flexible: The system doesn’t impose the route – if a driver decides to take another direction, it simply adapts and determines a new route to B.
- Support compliance: These systems are capable of advising against certain actions (for instance, driving wrongly into a one-way street) using signals and smart routing suggestions.
Does the driver feel he or she is part of a process? No, not really. That’s why we call the trend that we see in these solutions ‘No process’!
These ‘No Process’ solutions are proving their value in business as well. In a business context, they can be seen as flexible case management systems that have swarms of possible activities floating around customer cases. Based on technology for rules and events, for a specific customer case, the system can suggest possible ‘next best actions.’ This supports the employees to handle (navigate) the situations they encounter. In a way these solutions offer ‘just enough process,’ helping employees to pick the best (goal-driven) activity, but allowing the employee to choose another activity or path to deliver the best value.
No process - The perfect solution for today's knowledge worker times. Have a nice trip to your destination – happy customers and engaged employees!
Process driven transformation - exploring succesful change interventions(by Roeland Loggen)
Thursday, December 12, 2013
Sunday, December 01, 2013
Patterns of coordination (part I)
In my last blog item, I described a number of typical work patterns, that people usually pick when faced with the need to work together, in some type of (process) way. The main patterns that I will work in this blog item are:
- Procedure delivery: a pre-defined procedure can be established, which needs to be executed. The procedure might contain various decisions and paths, but all are known beforehand.
- Dynamic adaptive delivery: a set of typical activities (most standard) are known, but the path is emergent (due to unforeseen events, needs or insights). Many possible paths exist, too many to capture effectively in a procedure.
In this article, I will describe a number of typical coordination mechanisms that people pick how to coordinate the execution of the process, e.g. how are people triggered to get work done, and relate them back to the work patterns.
Pattern 1: 'Relay race' (Dutch: estafette)
This pattern can be used for procedural delivery, but this pattern is expensive, as the procedure is known but in every step, the coordinator needs to take action (e.g. many hand-overs)
The pattern is great for dynamic delivery, as the coordinator can assess the state of the case, and dynamically decide who should do what next to move forward in the case.
Pattern 3: 'Collaborative Planning & Monitoring'
In this pattern, a group of people, often multidisciplinary assesses a certain case, and together come to an agreed activity plan. The people will each perform their activities, and will update each other on the status and results.
This pattern can be used for procedure delivery, but again is expensive, as the procedure is known.
The pattern is great for dynamic execution, if the knowledge required to assess the situation and to decide on actions is advanced, and requires more people.
In a next article I will cover:
Pattern 4: 'Predefined Workflow Driven Orchestrator'
Pattern 7: 'Garbage can'
- Procedure delivery: a pre-defined procedure can be established, which needs to be executed. The procedure might contain various decisions and paths, but all are known beforehand.
- Dynamic adaptive delivery: a set of typical activities (most standard) are known, but the path is emergent (due to unforeseen events, needs or insights). Many possible paths exist, too many to capture effectively in a procedure.
In this article, I will describe a number of typical coordination mechanisms that people pick how to coordinate the execution of the process, e.g. how are people triggered to get work done, and relate them back to the work patterns.
Pattern 1: 'Relay race' (Dutch: estafette)
In the coordination pattern 'Relay race', there is someone that starts the process, following a certain trigger (an incoming e-mail, order form, etc). The person executes a number of tasks, and then hands over the execution to a next person, and so on.
This patterns works great for:
- Procedural delivery(see my previous blog-item)
Here, each person can find out the next in line, using the procedural agreements (process model / workinstructions)
- Dynamic delivery
Here, each person can decide, based on case specifics to whom he or she will hand over to.
Advantages of this pattern:
- There is always clear ownership
Disadvantages of this pattern:
- All activity usually takes place serially (where some parallel activities could speed up through put time)
- It takes time to find out 'where' a case is and what the status is
Pattern 2: 'Cloverleaf' (Dutch: klaverblad)
In the Cloverlead pattern, there is someone centrally coordinating the execution of the process. This coordinator or case manager will step by step perform certain tasks, then give an assignment to someone, wait, perform some tasks, and then again assign someone a certain activity, until the coordinator decides the work is done.This pattern can be used for procedural delivery, but this pattern is expensive, as the procedure is known but in every step, the coordinator needs to take action (e.g. many hand-overs)
The pattern is great for dynamic delivery, as the coordinator can assess the state of the case, and dynamically decide who should do what next to move forward in the case.
Advantages of this pattern:
- There is always clear ownership
- The coordinator knows what the status is
Disadvantages of this pattern:
- All activity usually takes place serially (where some parallel activities could speed up through put time)
In this pattern, a group of people, often multidisciplinary assesses a certain case, and together come to an agreed activity plan. The people will each perform their activities, and will update each other on the status and results.
This pattern can be used for procedure delivery, but again is expensive, as the procedure is known.
The pattern is great for dynamic execution, if the knowledge required to assess the situation and to decide on actions is advanced, and requires more people.
Advantages of this pattern:
- The team knows what the status is
- Progress is monitored and social controls can be used to adhere to rules and agreements
Disadvantages of this pattern:
- It requires quite some time of various people (cost, productivity) for the planning, status reporting and monitoring
- Disagreements might cause delays
- Unforeseen events require the response of the group, but people might not be (timely) available, causing delays or risks
Pattern 4: 'Predefined Workflow Driven Orchestrator'
Pattern 5: 'Situation Specific Configurable Orchestrator'
Pattern 6: 'Flexible Next Best Action Advisor'Pattern 7: 'Garbage can'
Subscribe to:
Posts (Atom)