LETTER TO THE EDITOR:

On the Definition of Project – In Response to Bill Duncan

9 March, 2009

Dear David,

Great to read opinion from the person responsible in PMI for the discussed definition of a project. That is what I expected – just comment from the author. Thanks to Mr. Duncan for his letter! But I will try to discuss with its explanations and answer questions – one by one.

ISO definition of a project is objective oriented and IPMA definition, while it is product oriented, does not require uniqueness of project products. This way both of these organizations avoided the trap of inconsistency of PM processes and project definition. The same with Prince 2, CMMI or BS 6079. My main critics of PMI project definition goes from the fact that this is internally inconsistent with the rest of PMBoK. The authors may define what they want and in a way which they choose but there is main requirement: it must be internally consistent. According to my research currently PMI with PMBoK is the only project management professional organization which cannot consistently define the basic concept of a project. So – sorry to say – I may not agree that PMI definition is as good as any other.

Please read carefully. I do not say that according to PMBoK it would be logical to know full specification of the product BEFORE the project starts. I say that – according to PMBoK – full product specification must be known DURING project chartering (or: starting). Because, again, according to PMBoK project definition, it is stated that only works producing UNIQUE products constitute a project. And works which do not constitute a project are outside of the scope of all the PMBoK – this is just PROJECT management body of knowledge and not

ANY_TYPE_OF_WORK body of knowledge. Please notice also that in PMBoK there is no place for making decisions about execution / cancelling after projects chartering (there is two step project initiation process for example in Prince 2). The decision made in Develop Project Charter Process is final and we do not have option for project termination at least until starting its execution. What is decided in Develop Project Charter process will be finally executed as a project. That is why people executing projects according to PMBoK must know whether a product is unique just in Develop Project Charter process.

I can see several problems with this explanation.

First – it looks like Mr. Duncan cannot see the difference between the very basic PMBoK (and, more general, project management concepts) like “product scope” and “project scope” (related to “project product” and “project work”). I try to discuss PRODUCT uniqueness (PMBoK project definition is just product-based) while he refers to project WORKS: “using different approach”, “different team” ("different location" will be discussed later) are concepts from the domain of project work and not project product. According to this argument laying egg by different hens would be close to the concept of project as each work is performed by “different (one hen) team”. And this is probably the very reason of the problem with PMI project definition – they mix project’s product (in concept definition) and project’s process (in concept explanation).

Second – the greatest usefulness of any definition is on “boundary” cases. The NASA Mars trip is obviously a project, also according to PMI definition. Perhaps every child in each elementary school all over the world knows that nobody visited Mars up to now. So we know about uniqueness of this product (and project) from the time when the idea was born. So in this case there is no risk of inproper classifying the works as "project". But what with less obvious cases? Like: is the work of building a house according to the same technical documentation in two different places a project in both cases? Of course it IS in each case, but this is questionable according to PMI definition. Of the differences cited by Mr. Duncan only one refers to product (and not to processes) – this is “different location” and it needs further considerations.

Third – we came to the concept of “uniqueness”. Does “different location” make a building unique? (I discussed this topic elsewhere but did not want to extend too much my initial letter to Editor of PM World Today). But if this is really needed, here you are. According to Cambridge Dictionary: (http://dictionary.cambridge.org/define.asp?key=86631&dict=CALD) “Unique: Being the only existing one of its type or, more generally, unusual or special in some way.” Do houses constructed according to the same documentation in different location conform to this definition? No. Does the process of producing a product in a different way (“different approach, different team”) influence the concept of this product (not: PROCESS) “uniqueness” anyway? No. So, applying Cambridge Dictionary definition of “uniqueness”, a house built according to the same documentation in different location may not be the result of a project. But for most of PM practitioner – it is.

Explanation of meaning of the word “unique” placed in PMBoK is rather unique.

No matter whether PMBoK offers a formal process for collecting requirements or not – they must be collected. The academic discussion whether this is a part of “product management” or “project management” is out of scope of our current discussion. The point is that in real life it often happens that detailed product requirements are collected, analyzed and defined only after chartering a project - and such a process is present in the current version of PMBoK. You must perform a lot of work after chartering a project in order to define project products. So you do not know what will be the project product when you charter a project and you cannot base project definition on product characteristics.

Let us recall my suggestion: "project is a temporary endeavor undertaken to satisfy internal business need or respond to external influence". What I have done was only collecting together some statements already present in PMBoK. So I understand that just the team producing PMBoK is the addressee of this comment. I do not know why Mr. Duncan could not see the addition of “a temporary endeavor” to the definition. Just bounding together temporary nature with business needs or external influence makes a project (both of these characteristics are taken form PMBoK).

I do not know what is meant by “building a car by Ford”. If Mr. Duncan thinks of building one car out of millions – this is not a project, as it is produced within a continuous process (does not conform to the phrase “temporary endeavor” in proposed definition). But when he thinks about preparing and implementing this car production process – this is a project according to proposed definition: temporary and performed according to external influence (customer demand). Processing a claim is just a process, which is not temporary.

Don’t we really know the goal of a project when we charter it? I ever discourage anybody from chartering a project until he / she knows its goal. And the last comment defending PMBoK, of a very logical nature: its authors placed “or” and not “exclusive or” in the cited text so you do not have to decide which option you apply in each case.

 

Best regards
Stanisław Gasik
Warsaw, Poland

sgasik@sybena.pl
www.sybena.pl

Back to Table of Contents

---