Having started to apply the profile successfully, I wanted to extend it to model measures,

These were modelled by meta-classing Class

As can be seen from this diagram, I have added a number of tagged values (which are modelled as attributes in the UML profile) to cover off the typical data that needs to be captured in relation to a measure.

In that post I described using Enterprise Architect’s ability to create custom UML profiles to create the beginnings of a custom modelling language for project benefits management.

In this article I walk through the basics of that UML profile.

Classes

The first task was to model the core objects of the benefits model – Objectives, Benefits, Disbenefits, Business Changes and Enablers.

These are all modelled as stereotypes of the Requirement metaclass:

The more observant of you will have noticed that there is also an enumeration called BenefitvalueType. this, together with the attributes on the Benefit and Disbenefit classes create a tagged value named “Value Type” in the final model, constrained to the different Sigma Value Types.

The Sigma Value Types are used as a way of classifying benefits -  this aids with the identification of measures, and also stimulates a conversation with stakeholders about missing benefits.

Relationships

The second part of the profile contains the relationships needed for the Benefits model. As I noted before, I have decided to use realisation and dependency links to model different aspects of the Benefits model. These are included in the profile by creating new classes of the same name that redefine the relevant metaclasses, thus picking up all the default behaviour.

The profile also includes a redefinition of Association, which I will use in the next part of the model, modelling Measures.

As always, I welcome comments!

The main sources in the literature I’m aware of are books by Bradley and Ward & Daniels. I’ve also had the privilege of learning directly from Gerald Bradley, so my own approach is very much influenced by his work.

A key tool is the use of visual maps, both interactively with stakeholders to discover benefits, and then as a way of presenting and communicating the complex causal links between an IT investment and the benefits it allegedly supports.

Interactive mapping works best with tactile materials – Post-It notes, sticky card etc. But for analysis and presentation some kind of tool is needed – drawing tools may work for smaller maps, but it very quickly becomes impractical, and something model-based is required.

Specialised tools are available, but they are just that, specialised tools: a good investment perhaps, but nevertheless a substantial outlay. The lack of affordable tools might, I suggest, be a block to wider adoption of these methods.

I’ve blogged before about using general purpose UML modelling tools to help programme shaping, so it was natural that I looked at extending this approach to benefits mapping.

An example benefits map using the UML approach is shown here, produced using Sparx Enterprise Architect:

I have created a UML Profile (which I will write more about later), which extends the Requirement metaclass provided in Enterprise Architect by stereotyping to create the five core Benefits Realisation Management objects:

Readers familiar with Benefits maps will have spotted something different about the arrows. Most graphical presentations use an arrow from the precursor enabler, change  or benefit to the subsequent change, benefit or objective:

Unfortunately this is not UML compliant, so  I have chosen to model using UML dependency and realisation relationships:

	Dependency	“This objective or benefit depends on that benefit”
	Realisation	“This change or enabler implements that change or benefit”

Using the language constructs in this way means that it is possible to use the traceability features within the tool to identify all the chains of dependencies.

Later posts will cover the development of the UML Profile, including the addition of attributes to the benefits and the modelling of measures.

I’m in the middle of a review cycle with a group of stakeholders who are used to talking about project benefits, but who perhaps have not used visual maps before – I shall blog how it goes!

What approaches have you used to document project benefits in a graphical format? Please leave a comment…

I’ve been familiar with the PRINCE2 method for a number of years, and have certainly applied the principles to local methods, but have resisted getting into it too formally because of the bureaucratic nightmare I have seen many organisations make from it.

One of the best things I can say about the Pearce Mayfield training is that through it I have seen, by contrast,  how to make PRINCE2 a living and breathing approach to delivering a project.

I’ve been “broadly familiar” with the PRINCE2 method for a number of years, but have resisted getting into it too deeply because of the bureaucratic nightmare I have seen many organisations make from it.

One of the best things I can say about the Pearce Mayfield training is that through it I have seen, by contrast,  how to make PRINCE2 a living and breathing approach to delivering a project.

I didn’t have time to attend the conference sessions, so this is somewhat less than a full review: I did however take part in one round-table session and had a chance to speak to a number of vendors at the exhibition.

“Simple, not Easy”, hosted by Adrian Dooley from The Projects Group was advertised as seeking “… to identify the fundamentals behind the evermore contrived solutions for consistently delivering successful projects”.

I’m not sure we achieved that in 90 minutes, crammed into a room that showed an artful ability on behalf of the hotel to turn a fire escape corridor into a meeting space (!!), but there was some entertaining discussion, capped by Adrian’s takeaway – “Practice what you preach” – i.e. as professionals in achieving business change it’s down to all of us to address the business changes needed to make project management more effective in the organisations we work with.

There was a small exhibition area with a number of vendors of software, consultancy and training. Ones which caught my eye were:

• Change Director
• PAT from NinthWave
• RiskAid from Risk Reasoning

I’d have liked to take a look at Simply Project Office from Mundane Software, but sadly they seemed to suffer from the mundane problem of having no-one on the stand when I looked!

Proactive Risk Management, Controlling Uncertainty in Product Development
by Preston G. Smith and Guy M. Merritt

In Lean Production there exists the concept of the “visual workplace”, commonly expressed through the 5S model. Hal points out that projects may not always involve material products and resources but always involve people and conversations; it therefore makes sense to translate the 5S model into what he calls the 5R Protocol for a Listening Workplace:

1. Roles
2. Rules
3. Reflection
4. Relationships
5. Routines

What’s interesting is the way his own thinking is developing as he reflects on this model and the conditions that need to be in place for real changes to happen – critically the need for having the right mental distinctions to notice what is really important and then taking action based on those distinctions:

What we notice has to do with the distinctions we can make and the routines that we follow. Both our noticing and effectiveness in action increase as we take action. If we want to work in a lean way we need the distinctions of lean and we need to take action. [...] Learning to operate in a lean way happens by doing projects in a lean way.

For me this sits well with the model of cognition used by NLP:

Our habitual perceptual filters control what we actually notice in our surroundings – an engineer will notice different things from an HR expert. The mental programs we use (or habitual ways of thinking) will then influence what meaning we ascribe to those things and therefore influence our conscious intent about what to do. Those same mental programs will distort our conscious intent into our everyday strategies, which in turn result in actions and words that fit with our perceptual filters. The whole system is both recursive and self-reinforcing – the success of actions we take in the world tends to strengthen the perceptual filters and mental programs that led to us choosing those actions.

In such a model changing behaviour often needs the conscious adoption of new filters and disctinctions re-inforced by action until new unconscious mental programs take hold. This is where coaching is especially useful to remind the person who is changing what they should be paying attention to.

What Hal is doing with his 5R model is start to express the things that make a difference in order to get “Lean Projects” right – it will be interesting to see how he develops this into practical tools that can not only be applied but through their application embed new ways of thinking.

What I love about the project management link is the way the functionality of this tool complements the total left-brain-ness of standard project management tools. The most important part of any project is the first meeting where the people involved get engaged with breaking the scope down into manageable chunks – to be able to do that with a mindmapping tool and then export a first-cut WBS or PBS is just…cool…

I’ve played around a bit with mind-mapping for a number of years and although you can’t beat the flexibility of pen and paper for personal notes, for collaborative work some kind of electronic tool seems essential. (apart from anything else, I lose paper!)

Links to old broken wiki removed
The other idea that’s nagging me tonight is a need for a MMToWiki tool. I’ve slowly started putting some NLP Wiki pages together but I’m finding the flat-file format of a Wiki rather frustrating when writing a set of interlinked documents. I’d love to be able to outline and write the first major tranche of those pages in MindManager, then export to a set of Wiki-formatted text files.

Hmmm… and when, I wonder, is tool-building a displacement activity from the writing?

[update 2003-02-13] Have found Mind2XML an add-in for MindManager that does “exactly what it says on the tin”. So now the gap in my knowledge that comes into focus is how little I know about XSL

[update 2007--4-11] And almost exactly 4 years after I wrote this post, a very similar idea about the combined use of wiki and mindmaps emerges over at Activityowner.Com – with the big difference that he has actually produced a first draft of a conversion tool…