(R)Evolution in Project Data
How should we best adopt project data analytics? It could revolutionise the delivery of projects, but a new study by the University of Warwick identifies the barriers that prevent organisations reaping the potential rewards. Professor Naomi Brookes and researchers Luis Lattuf Flores and Elaine Falconer explain…
It is ironic that, despite having more data than ever before on our projects and a plethora of digital tools to hand, there is little evidence to suggest any marked improvement to decision‑making. Initial studies have demonstrated that while project dashboards are widely used, they have not yielded the improvements in decision‑making that would be expected. A key challenge is that dashboards generally rely on past performance. But, as anyone who has dabbled in the financial investments market can attest, past performance is no guarantee of future results. Data is also greatly simplified, reducing the complex interdependencies into a set of colourful charts.
But things are about to change. We are on the cusp of a project data revolution where every morsel of project data produced will be fed into the decision‑making process through the establishment of project data analytics (PDA). Given the huge potential for PDA to improve project delivery and its failure so far to live up to its promise, there is a pressing need to understand the factors that are impeding the adoption and the enablers that will help us to overcome these barriers.
The Project Praxis research group, based at the University of Warwick, has been working with a number of major infrastructure delivery organisations to understand these barriers to identify practical ways to overcome them. Its most recent study, sponsored by the Oakland Group, engaged with senior business managers to understand their experiences with data analytics and capture their views on what they felt were the barriers and enablers to widespread PDA application. This provided a unique insight into the minds of the key decision‑makers and the opportunity to understand which direction we need to move in to harness the transcendent decision‑making power that comes from analysing all our project data at the same time.
Take action now to ensure good data quality
The first theme identified by the participants related to the challenges associated with a lack of consistency across organisational units because of the historical evolution of management systems. It was found that projects often encountered difficulties as a result of basic failures to use common definitions throughout their organisations for items as fundamental as project identifiers. The solution to these challenges was clear. There is a pressing need to bring about the standardisation of data management systems and processes so that consistency in data collection and management is achieved across all operating units and projects.
It was also clear that action can and should be taken now, as good‑quality data is a fundamental prerequisite to any future analytical endeavours. As we were told by one senior project controls executive: “The underlying issue… isn’t that there are many systems, it is that they all may be coded differently.” A senior project management office executive in transport infrastructure explained that: “The problem we have is that data formats are not compatible a lot of the time; it would really be a step forward just agreeing what data format we will all use; for example, coming up with a standardised data format or data manual.”
Cut through the jargon
Participants were very clear that they did not feel the need to have cohorts of data analysts joining the ranks of project practitioners. What they required instead was to develop a ‘data literacy’ across all members of the organisation related to what data is and how it can be used. A key enabler was found to be cutting through the jargon around topics like PDA to give people a basic understanding of what PDA can deliver and how it might contribute to decision‑making.
One senior project controls executive told us that: “There are a lot of a variable skills across our organisations in analysis, understanding, interpreting of information and displaying of information, and that’s even before we get into really complicated statistical analysis. Therefore, we have some real basics that are missing, so it is worth discussing and understanding the value of establishing centralised methods of control of data versus widespread organisational competence and understanding of these principles.”
Another senior data and analytics executive explained that: “There are so many buzzwords out there. I have a lot of customers that will talk about wanting to do machine learning… but when you start to drill into it, you realise that they just picked up the buzzwords without actually understanding what they mean.”
Stop trying to eat an elephant
One of the major barriers in terms of implementing PDA was found to be the sheer magnitude of the implementation task, requiring the simultaneous consideration of the organisational culture, current data management processes and skills gaps that will need to be closed to enable successful implementation. It is unsurprising that participants identified consultants as a potential enabler for the smoother introduction of data analytics. However, participants also identified that the use of outside organisations could be a barrier to developing and embedding the much‑needed in‑house ‘data literacy’.
“It’s a long journey isn’t it? How do you get going on it? You are competing against other transformation initiatives. Trying to eat the elephant all at once is too hard, too big and too culturally challenging to make a shift,” one commercial director at a data consultancy told us. A senior project controls executive also shared that: “If we continue to rely only on external organisations, we continue to sort of rob our own internal organisations of the chance of potentially developing some of those skill sets. So, it’s really worth trying to balance that and really build some internal competence and data literacy.”
Seeing is believing
Participants from project service organisations identified that client organisations were unwilling to fund untested, intangible data analytics activities, while client organisations identified that executive decision‑makers are often more prepared to trust their instincts than rely on what is perceived as subjective data analysis. However, it was widely agreed that providing clear evidence of successful PDA applications is fundamental to securing investment in the technology and wider acceptance from senior budget holders.
Professional organisations, such as APM, were seen as key enablers in uncovering and showcasing the pockets of current examples of PDA and celebrating successful applications in this area. “Seeing is believing, and by doing those little pockets of PDA projects within a company, it starts opening up those opportunities,” explained a senior data and analytics executive.
The truth will set you free
It is ironic that a profession dedicated to implementing change is seemingly averse to change. However, fear of the unknown was identified as a major obstruction to PDA implementation. This fear was found to be manifested in the perception that PDA could potentially take the control of disseminating performance data out of the hands of project deliverers, thereby removing the ability to manipulate project data to ‘soften the blow’ of bad news related to poor project performance.
It was also acknowledged that fear needs to be addressed around the idea that artificial intelligence, and by association PDA, will result in an elimination of certain roles in project delivery and substantively change the remaining roles. One of the crucial ways to overcome this aspect of fear is to increase project professionals’ data literacy and to take learning from other industries that have successfully implemented and benefited from data analytics. After all, data analytics is not new and, in reality, project management is a late adopter of the technology, so there is a huge amount of learning and experience that can be taken from other industries.
“People see things emerging and ask, ‘What does it mean for me?’ There is a general reluctance to adopt predictive analytics only because it is out of their comfort zone,” one senior assurance executive told us. Isn’t it time to take that courageous first step?
The authors would like to acknowledge the invaluable support of the Oakland Group in undertaking this research. The Oakland Group is a world-leading data consultancy operating at the intersection of process, analytics and governance.
What is predictive data analytics?
Adopting PDA successfully goes way beyond the simple acquisition of a new piece of analytics software or a new dashboard. At its simplest, it is the use of past and current project data to enable effective decisions on project delivery. This includes descriptive analytics, which presents data in the most effective format, such as in a dashboard, and predictive analytics, which applies machine learning to a multitude of data sources to predict future performance. Machine learning involves computer programs that spot patterns between characteristics of projects or programmes and performance. This process gets more accurate the more it is used.
The top five actions to break down the barriers to successful PDA implementation
1. Improve data quality
The prerequisite for successful PDA is good‑quality data, and this is dependent on how mature the organisation is in terms of data management. Organisations need to first understand how they currently collect and manage data by assessing the protocols and tools that are used, the skills of individuals across the organisation and the motivation and culture within the organisation to adopt PDA.
2. Create awareness not analysts
The management of data needs to be the responsibility of everyone and not seen as a specialist role. Organisations need to identify how they can develop their ‘data literacy’ whereby all members of the organisation have an understanding of the role of data and how it might contribute to decision‑making.
3. Manage the size of the change task
Organisations adopting PDA need to realise that there is no silver bullet for implementation. Embedding PDA will require prolonged commitment and needs to be adequately funded and resourced. Outsourcing tasks and engaging with specialist consultants may overcome current skills gaps in this area. However, to truly embed PDA into the organisation, investment into developing in‑house capability is required.
4. Share examples of PDA on live projects
There is a pressing need for organisations to showcase examples of PDA being successfully applied on live projects to develop a greater awareness and appreciation of how PDA can be applied to projects and demonstrate that there are tangible benefits to the approach.
5. Eliminate fear from project delivery
Fearing the loss of control in how data is shared and interpreted was identified as a major barrier to implementing PDA. However, taking learning from other industries that have successfully used the technology will help overcome this fear.
0 comments
Log in to post a comment, or create an account if you don't have one already.