Delivering successful technology projects is hard and many, particularly in the public sector, do not hit their deadlines, stay within budget, or achieve their objectives.
QAO found that of the projects on the dashboard, only 47% are tracking on time, with 26% expected to take at least 50% longer to complete than expected. 23% of projects are expected to cost 50% more than anticipated.
The Queensland Government is not unique in struggling with the delivery of technology projects. A 2018 survey of 3,000 public officials from 18 countries, the McKinsey Centre of Government found that 80% of public sector transformations fail to meet their objectives.
The QAO report outlined five factors that contribute to successful technology projects:
- Senior leaders actively lead and challenge
- Projects are aligned to business outcomes
- Internal and external teams work towards the same goals
- The team has the skills and capacity to match the challenge
- Learnings are identified and acted on.
We agree that these factors are important to delivering successful projects.
However, in our experience, most projects will already be aiming to do these things. Below we have outlined some additional perspectives garnered through many years of working on public sector projects.
Transformational projects are challenging. Not only are they often large and complex, but people find it difficult to describe systems and processes they have not seen, and it is hard to know what people will want or need in the future. Eliciting this vision – and then communicating it to a vendor, with an effective commercial contract behind it – is a significant challenge for nearly all organisations (not just public sector).
Nearly all projects start out optimistically. Typically, officials and project teams have every intention and expectation that the required service or solution will be delivered on time and on budget. This is due to the inherent optimism bias that we all share. However, these biases and assumptions often contribute to the opposite result: over time, over budget, and a solution that is well short of expectation.
This optimism bias, combined with the sunk cost fallacy (a belief that a committed, unrecoverable cost is relevant to future decisions), too often results in a hesitancy to make a tough call to modify or close an underperforming project until well after it is needed.
So, what to do?
In our experience, increasing the chances of success relies on four aspects.
Firstly, scope and requirements must be managed rigorously. The idea that rapid development methods (e.g. Agile) obviate this requirement, or are even necessarily applicable to many large public sector projects, is a fallacy. And we have seen too many projects fail because the rate of scope creep (changing customer requirements) exceeds the vendor’s capacity (or commercial freedom) to respond.
Secondly, avoid ‘big bang’ delivery! In other words, don’t try to deliver too much all at once. Big bang is a highly reliable indicator of a project destined to run into trouble. Queensland’s Chief Customer and Digital Officer captured this point in his own response to the Acting Auditor-General: “…simplify and shorten implementation cycles that deliver outcomes continuously … smaller, cheaper projects, more often.”
Thirdly, get the contract right, and use it. You need the right contract for the type of project (e.g. software/technology development, customisation, as-a-service) and must be prepared to amend it if necessary. Legal teams need access to expert independent advice to get the language right and embed reality. The contract must provide the foundation for success, throughout the process.
Finally, nothing beats a critical friend who tells you how it is. External assurance should be just that – if you aren’t getting some uncomfortable advice, there’s probably room for improvement. In our experience, genuinely independent external assurance is a reliable way of improving project decision-making and improving outcomes for the public.
Written by GWI CEO, Neil Makepeace & GWI Partner, Vanessa Douglas-Savage
