Photo courtesy of Brian Wernham
Agile Profile is a regular feature profiling leaders in agile government.
Who
Brian Wernham
Author, “Agile Project Management for Government”
What is your definition of Agile Governance?
There has been a great deal of work on Agile methods at the developer/team level, but little debate or discussion on either the governance of agile projects, how to govern ‘hybrid’ programmes that have a mix of agile and waterfall sub-projects, nor at the Main Board’s level, applying ‘agile’ concepts to a portfolio or change. The £600m Universal Credit programme in the UK is an example of a major organisation not having a framework within which to govern a major programme using Agile methods.
What are the advantages?
The UK Association for Project Management has decided to produce a (slim) guide on ‘Agile Governance.’ Note the word ‘slim!’
The guide will focus on behaviours, rather than processes. The format will be similar to the previous APM ‘governance’ guides (glossy A5 and 12-18 pages long). The guide is being developed under the auspices of the ‘Governance’ Specific Interest Group (GovSIG) of the APM which focuses on CxO/Main Board behaviours. It will take the perspective of innovation change initiatives where a traditional up-front planning approach may not be flexible enough.
Agile behaviours are becoming increasingly important at enterprise level. Organizations are starting to recognise the potential for innovation to disrupt their status quo … something that is best seen as an opportunity rather than a threat. Adaptation and emergence is required in the face of rapidly changing requirements in technology and business, with drivers such as time to market and finite budgetary constraints. Above all, there is the need to consider validated learning when formulating a strategic response.
We are concentrating on the essential behaviours that, if propagated from the generalists on the Main Board through to the most detailed technical specialist, will increase the rate of effective application of innovation, and reduce the risk of inappropriately over-planned and stage-gated projects that often fail in a misguided ‘big-bang’ attempt at enterprise change.
What are the problems?
The watchword here is ‘incrementality.’ The Main Board must ensure that the portfolio of change initiatives is measured on practical feedback from the ‘coal-face’ – from the expert community who are beneficiary (or business user), business receiving benefit (economic buyer) to ensure that investments are on track. Actionable rather than vanity metrics are to be favoured, and reliance should not be placed on theoretical strategies and unproven plans. Development money that is tied up as work-in-progress, yet to be implemented and subject to depreciation until it is used, is reduced. These behaviours need to be exhibited at every level – from Boardroom to every project member. Organisational gravity is the enemy and executive sponsorship for agile change will be instrumental in overcoming it.
What do the feedback loops look like in ’Agile Governance?’
The focus is on being incremental and delegating every decision to the lowest possible level. If ‘just enough’ planning is carried out, and multiple paths are explored and kept open until the ’latest responsible moment,’ then top management will create an agile strategy and culture.
The focus for innovation must be on business priorities, and the governance approach need to know when to start and know when to stop. The feedback is by measuring progress using agile metrics, such as value delivered per sprint, not by focussing on the cost of each task carried out.
Objectives must be prioritised, even when they all appear to be equally essential.
Where do you set the limit for project control for Agile governance? Portofolio, programme or projects? And why?
Agile governance starts at Main Board level with the whole portfolio of change.
In the past, this has targeted maximum return on capital, often underestimating the increased risk of large projects and the delay to start of benefits of over-long implementation. The focus of an Agile portfolio is to reduce work-in-progress and increase flow of work. To do this, the portfolio director needs to re-shape innovation projects to deliver incrementally and to eliminate stop-start decision gates. This may be contrary to traditional approach of having a few harsh decision points. But it makes more sense: have many decision points – and make them all based on assessing ACTUAL delivery rather than assessing theoretical specifications, estimates and plans.
In what way is the fluid nature of agile thinking an advantage in governance?
For example: we know from lean manufacturing that if we reduce batch sizes we reduce excessive work in progress. Procurement ’optimisation’ often creates unconstrained demand for larger and larger batches of work which promise slight cost savings through economies of scale, but never get completed.
Can you run Agile Governence on waterfall project? How do you do that?
The APM Webinar I led on September 11, 2014, will be available on the APM YouTube channel shortly.
In that Webinar I interviewed Karen Richey of the US Government Accountability Office (G.A.O.). She explained how an overall ’Master Schedule’ can be planned using a waterfall approach, within which iterations of Agile activity can take place – each Sprint towards a release not being planned in detail with a traditonal Gantt chart, but instead using a Prioritised Backlog/Requirements List. The MoSCoW approach gives priotitisation: with ’Must-Haves’, ’Should-Haves’, ’Could-Haves’ and a list of ’Won’t-Haves’ which are outside the project scope. The team work consistently to a timescale down the list of priorities, and ALWAYS delivery something of value at regular intervals. These releases can form part of a waterfall Gantt chart.
What will be next? If agile PM is a way to deal with quick external changes that the classic Waterfall couldn’t handle, What will come next?
The future is Evolutionary Algorithmic Development – where many innovations are developed simultaneously using automated approaches, and they tested in real-world simulations, with only the fittest surviving.
Leave a Reply