Edited transcript from the Online Workshop held here on 15 May 2000 on Process Improvement methods, chaired by Robin Cox.

Robin Cox: I'll start by explaining that I raised Process Improvement as a subject for the workshop, having got into the subject in 1997 when I was working for a large corporate in the software industry. I ended up running about 40 workshops for them. However, I do not present myself here as a guru by any means. My aim is to encourage some sharing of experience, as I believe PI is everyone's business. Could you say what your interest is?

Laura Lindell: Okay, I have limited experience in delivering process review or improvement programmes but have worked as an assoc and delivered a programme to a major airline. The day was not well received and I simply wanted some ideas n the key principles of PI

Tim Pickles: My own knowledge is more limited. A couple of colleagues are involved in PI as part of their organisational development consultancy work - I'd like to understand more of the techniques involved.

Robin Cox: I suggest we look at PI first in order to help people to understand what a process is. What is your experience of doing this?

Andy Waterworth: I have done quite a bit of work on this as I used to work as a training manager for a manufacturing group

Robin Cox: Andy, how did you clarify what a process is, then?

Andy Waterworth: In simple terms it was easy to do in a manufacturing environment, we classed a process as anything that happened to the product as it moved through the site. these processes were classified as Macro and micro processes.

Laura Lindell: So is it anything which transforms your input into an output?

Tim Pickles: In those organisations where I've come across the term, it was regarded as a mapping exercise - to define the process for delivering a service, etc and then identify restrictions, feedback loops, etc.

Robin Cox: In the software context, I used the example of a design review, in which a draft design specification is debugged. But in truth some other examples might include cooking a meal, or playing a piece of music.

Robin Cox: Laura, yes, plus: the resources to do it, the procedures you follow and the policy that backs them up.

Laura Lindell: I assume that we are more interested in the controls that are applied at each stage of the 'process'

Robin Cox: The important point is that a process is what you do, and the map is the process description.

Andy Waterworth: Yes Robin, one of the exercises I used to run involved boxes of matches, the process the learners had to create was how to put matches into the box and stack the boxes. sounds simple until you escalate the matches to hundreds of boxes

Tim Pickles: What sort of tools are you using to identify and define the process - its sounding as though very graphical tools are most useful

Robin Cox: Yes, I use a flow charting aid such as Visio. Incidentally, there's quite a good intro to this at http://deming.eng.clemson.edu

Andy Waterworth: I have also used visio and haven't found much that is better.

Laura Lindell: When I did this I had the group define the process of airport check in - problem was we didn't have a clear definition of the desired output. At what point was the process complete.

Andy Waterworth: what happened Laura

Laura Lindell: Yes, tols help but only if you are clear about what you re transforming into what. When does the specific process that you are measuring start and finish

Robin Cox: That's a good point - the process description needs to say what state the inputs and outputs should be in, e.g onions chopped or sliced!

Andy Waterworth: I guess that it why most process improvement programmes begin because the inputs and outputs are unclear

Laura Lindell: What happened would take too long to describe but I was working with a group who all were stakeholders in the process and who had different ideas on the start and end points. e.g. is a customer done when they are in the lounge or actually on the aircraft?

Robin Cox: So we really need some criteria by which to judge a decent (i.e. workable process).

Laura Lindell: Sounds like a good start

Laura Lindell: There must be processes within processes

Robin Cox: How about these? A workable process is practised, enforced, trained, documented, measured, and capable of improvement. This comes from the Software Engineering Institute at www.sei.cmu.edu

Tim Pickles: I can see the desirability of agreed inputs and outputs (and the problems which arise when the parties are not agreed) - but I was thinking that PI is more about analysing the processes themselves, rather than reaching consensus on inputs and outputs - which is more of a goal-setting task.

Andy Waterworth: This was why we classified as Macro and Micro processes, so at a Macro level it may be Check in desk to lounge to Aircraft. Within the Macro processes are Micro i.e. check bags get check passport!

Tim Pickles: And the problem in so many organisations is that people don't have a shared, agreed definition of what they're working to achieve !!

Robin Cox: Hence the criterion that a decent process be documented.

Laura Lindell: So we are back to the question of what tools assist this process

Robin Cox: Tim's addition of "shared" to the list of criteria is important. With regard to tools, we need to think about measurement tools, and team building aids.

Andy Waterworth: The best tool to assist the process has got to be the people who are performing the task. Ownership has to be one of the keystones of Process improvement

Tim Pickles: And is the value of PI proportional to the size of the organisation - larger organisations require more process documentation as part of 'knowledge capture' (i.e. less dependence on the presence of key knowledge-holders)

Robin Cox: I wonder if small organisations don't also need to capture process knowledge, as the impact of someone leaving is proportionately greater?

Tim Pickles: Yes, good point

Robin Cox: Laura, what is your experience in PI? You said it didn't go well for you? Can we learn from this?

Laura Lindell: Yes, but recording a process is more likely to be necessary where there are multiple departments or stakeholders involved. This should in principle assist the review

Robin Cox: I had problems running workshops in PI initially, as I was confronted by staff who felt that they did not have to consider processes, as long as they followed procedures.

Laura Lindell: This isn't one of my specialist areas. I was simply hoping to eavesdrop to pick up a few tips. I had a day with a group of very senior managers who wanted to learn how to review their processes. After defining a process they experienced trying a few different 'tools' as selected by the organisation. these tools included 'Swim Lanes' Force Field Analysis and visioning. Are you any wiser? They weren't.

Andy Waterworth: force field analysis and visioning are probably the first step in process improvement with groups as they scope out what the end product should look like

Robin Cox: I think that the point here is that PI focuses many if not most of our work-related training methods in such a way that we can actually envision and measure the benefit.

Tim Pickles: But I can see where the confusion comes between 'procedure' and 'process'. How would you define the difference to a group, Robin?

Andy Waterworth: With process Improvement the wide scope or vision is OK but the "devil is in the detail" the glaring gaps of 19th century factory are gone. Most businesses these days are looking to shave of seconds and minutes, not hours and days, or save 0.2p over a million units. This is only possible with detailed process, identifying the space for change.

Laura Lindell: In training it is probably easier to take a group of people who have a common process that they have an interest in reviewing. My group were all heads of different areas simply wanting to experience some new tools

Robin Cox: Process includes procedure, but it also includes the resources with which to implement the procedure, and the policy or rationale for the procedure.

Tim Pickles: So Process is a bigger-picture approach encompassing all elements of a situation - and ensuring they are in balance ...

Tim Pickles: Robin - were coming to the end of the time for this session - thanks very much for your input to it - a small, but quality group! - are there some final thoughts and ideas about PI arising from this discussion?

Laura Lindell: Thanks for your time guys - I've had time to 'define, analyse and improve' my model.

Tim Pickles: OK folks, we'll post an edited version of the transcript in the next few hours, for reference

Andy Waterworth: Thanks all I Haven't talked about process for a while I miss the smell of oil and the crunch of swarf under my feet,