Is the Life Expectancy of a PMO Really Only 2 Years?

 

 

 

 

 

 

 

 

In his recent PMO Newsletter, Americo Pinto took on the problematic and oft-repeated claim that “PMOs have an average lifespan of 2 years.”
His focus was not only on challenging the statement itself and how it’s been misused or misquoted, but also examining the context it was made in, and what the real intent of authors was when it was made.

As a reminder, the statistic/statement most likely comes for a 2010 paper by Monique Aubry, Brian Hobbs, Ralf Müller, and Tomas Blomquist (neither Americo or I could find any other sources) and says:

Empirical evidence shows that PMO life expectancy is approximately two years.

First, let’s start with a bit of context, and include the entirety of the thought –

Despite the dynamic context within which PMOs try to solve organizational issues, a mistaken paradigm exists that changes in PMOs, or even termination, are due to *wrong* configurations of PMOs, which need to be corrected through transition into new and *good* PMOs that will last much longer that their former versions. However, this long-term effect rarely takes place. Empirical evidence shows that PMO life expectancy is approximately two years. PMOs are dynamic entities created to solve specific issues within dynamic organizations. That implies a short life expectancy for PMOs.

In other words, the short life expectancy isn’t a bug, it’s a feature. Many PMOs are (or were) created to solve a problem, they solve it and then are shut down. And it makes sense – PMOs cost money, so if the PMO has solved the problem it was created for why would you keep spending money on it?

And that’s the problem here – the 2-year claim is almost always used as an indictment of the ‘value’ of PMOs. Which, as we’ll see as go further, isn’t the case at all.

Now, Americo has done a fantastic job of addressing the ‘intent’ behind the paper (and I really have nothing to add), so I’m going to take a slightly different approach and focus on just one aspect of the statement – ‘empirical evidence shows…”

Because ‘empirical evidence’ means it must be true, right?

Right?

In their paper Aubry & Hobbs include three citations to support this statement (the ’empirical evidence’) – Hobbs & Aubry, 2007 (one of their earlier studies); Interthink Consulting, 2002; Stanleigh, 2005.
I’m going to address these in descending order of age.

Before we start, a quick note about the studies/surveys –
Both papers by Hobbs & Aubry were academic undertakings and appeared in peer-reviewed journals. So, we can be reasonably confident in the methodology they used for their work. The other two, Interthink and Stanleigh, were surveys done by consulting companies, which means we don’t have the same insight into the methodology or specifics of the studies. So, in many ways the results or ‘evidence’ of these two are open to the same types of critique/criticism the Standish Group’s CHAOS Report has so often endured – population, survey structure, definitions, bias, and so on. Fortunately, for our purposes these questions aren’t really a factor, as you’ll see.

Okay, back to the ‘empirical evidence’….

“Hobbs & Aubry, 2007” refers to their paper titled “A multi-phase research program investigating project management offices (PMOs): The results of phase 1”. This paper included survey responses from 500 PMOs, from what appears to be primarily North America & Europe.

In this paper they make the following two statements:

PMOs have been popular since the mid- to late- 1990s. Surprisingly, 54% of the of PMOs in existence today were created in the last two years, according to data from 2005.

and

Two phenomena are at work producing this result. First, new PMOs are being created at a relatively high rate. Second, PMOs are being shut down or radically reconfigured at almost as fast a rate. The result is a population dominated by PMOs that have only been in existence in their present form for a few years…

Now, there are a few important things to note in these sentences.

1. This is all that was written regarding the lifespan/expectancy of PMOs in the paper.
2. The ‘two years’ indicated their current age, not how long they lasted.
3. The 2nd phenomenon mentioned points to ‘shut-down’ OR ‘reconfiguration’, but there is no indication of how many or what percentage were shut-down vs reconfigured. However, the last sentence suggests that the majority were likely reconfigured vs shut down with their emphasis on “population dominated by…in their present form” vs new. Americo addressed this aspect in greater detail in his newsletter as well.

So, we have 46% of the 500 PMOs surveyed being older than 2 years, and the rest either having been created within the past 2 years (so simply ‘young’), or reconfigured (which starts the ‘life expectancy count’ over). But nothing to suggest, or support the ‘life expectancy’, or ‘birth-to-death’ of a PMO is ‘only’ 2 years.

Next, the Stanleigh paper.

In 2005 Michael Stanleigh, through his company Business Improvement Architects, conducted a survey of “750 organizations around the world”.

Now, the results of this survey are a bit more damning.

Many organizations set up a Project Management Office (PMO) as a tool to improve project management. However, the evidence shows that these fail more often than they succeed. In fact, the research indicated that over 75 per cent of organizations that set up a PMO shut it down within three years because it didn’t demonstrate any added value.

This certainly doesn’t look good for PMOs, their life expectancy, or their ‘value’.

We don’t know exactly how many of the 750 orgs studied had PMOs (his survey wasn’t specifically on PMOs), but worst case is 73% of 750 means 548 PMOs lasted “less than 3 years”.
This is both empirical evidence of the life expectancy claim, and an indictment of the value of most PMOs.

Not good for ‘Team PMO’.

BUT, if we read a bit further we find some hope –

In fact, 80 per cent of organizations in the research study had no formal business case for the development of their PMO, while 73 per cent identified “lack of executive sponsorship” as the primary reason for failure of their PMO.

So, of those 548 PMOs that lasted less than 3 years, 438 had no formal business case when they were created, and 400 had no executive support after they were established.

Now, 75% of PMOs shutting down within 3-years isn’t a number we can simply disregard and definitely qualifies as ‘empirical evidence’, But the lack of a formal business case and executive support does raise some ‘correlation/causation’ questions regarding this number. If a department is created and then shortly thereafter closed down, and the cause is due to lack of support or a formal business case (i.e. leadership failure), should those numbers be included alongside the successful, reconfigured, or young PMOs to calculate the ‘average life expectancy’? They certainly skew the average down, but he questions is, do they do it fairly?

Lastly, the Interthink Consulting survey.

This survey was conducted in 2002 and included 181 participants representing 78 organizations.
In the section of the Executive Summary titled ‘PMO Duration’ the survey states the following:

In looking at the longevity of PMO capabilities, nearly half of all respondents indicated that their PMO had been in place for less than 2 years, reflecting the relatively new recent focus that this been placed on establishing organizational project management capabilities. Given this high percentage of relatively new PMOs, it is perhaps more surprising that 16% of PMOs have been in place for at least 5 years, with nearly six-tenths of those (10% of respondents) having been established more than 10 years ago.

So again, it’s not really saying PMOs ‘only live’ 2 years, but rather that the majority of those surveyed are only 2 years old ‘so far’.

After all this we’re left with several take-aways –

1. The intent of the ‘empirical evidence’ statement wasn’t to suggest the majority of PMOs ‘only’ live 2 years. Two of the three surveys highlight this number to indicate the rapid implementation and adoption of PMOs.
2. ‘Life expectancy’ was a poor word choice. It suggests ‘beginning to end’ and doesn’t account for ‘so far’ or ‘since it changed’.
3. These studies were written in the early days of PMOs, so the 2-years figure is really talking about their ‘youth’, not their lifespan. In many ways this would be like looking at a kindergarten class and concluding the ‘average life expectancy of humans is 5-years’.
4. We really have no idea what the average life expectancy number is. In fact we have no real idea of what the average age of current PMOS is. This small set of comments seems to the both the start and end of this line of research, as no other research since has sought to confirm, challenge, or even update this number.
5. As Americo’s post suggests and reinforces, 2 years is a ‘starting point’ and begins the change and evolution process. It’s not an end point. 5. This 2-year idea, and all the supporting data is 15-23 years old, and each of the papers included cited how early this was and the rate at which PMOs were being created. If the number of PMOs was on the rise 15 years ago, then we can only guess at how things look now.
6. Most importantly, as might be inferred from the Stanleigh paper, you need a business plan and executive support for your PMO to even have a change to survive/succeed.

And ultimately, the claim that ‘the average life expectancy of PMOs is 2 years’ is badly misunderstood and misused, and needs to stop being repeated.

 

 

Originally published on LinkedIn