In the second part of our extensive interview with Niall Barry, the head of IT at the Irish Government’s Department of Social Protection (DSP), we learn about evolutionary approaches to large-scale change, and why conventional wisdom about public-sector IT is often wrong.
When you’re embarking on a major business and technology transformation, how did you go about making that change happen and also ensuring buy-in from all staff?
I think that effective transformation essentially relies on convincing the business that new approaches actually work and will bring benefits at least commensurate with the investment you are asking them to make. To bring in a whole new approach, we needed to convince senior management that the ‘theory’ was sound and then needed to convince people ‘on the ground’ that the theory worked in practice.
I’ve already mentioned the need to find champions in the business who are willing to listen, try to understand and work with IS [information systems team] to achieve the benefits and that applies at both senior and operational levels. There’re very few people who don’t want to do their job better and I have to say that people at all levels were willing to be convinced. Having said that, there is still a natural reluctance to move on from existing systems and practices.
It is extremely important to involve staff and to show progress early and often so that we can build on successes (and deal with setbacks) as quickly as possible. This means that the Lean and Agile approaches are a very good fit. When staff collaborate on developing a prototype, they are more inclined to buy into the overall result and it is more likely to solve real business issues. As we progressed to other business areas, we were able to point at the work done so far as evidence that the approach worked.
You chose an Agile methodology: how does this work in practice for DSP, and what were the benefits?
‘Agile’ is fundamental to the IS development process and mirrors the Lean/Six Sigma approach in the business. I think that ‘agile’ is a natural fit with, and is facilitated by, object-oriented development. I’ve mentioned that the public-service culture is one of ‘accuracy and completeness’ and this means that there is a fundamental tension with any approach that focuses on core requirements first and delivers on an incremental basis.
I don’t believe that you can just mandate Agile or Lean, though. I believe that you need to buy into them as a concept and then implement them in an agile way; mainly pick the bits that are more likely to work in your environment, implement them, improve them and consider other bits and supplementary approaches as you gain experience. It took a bit of effort to get started but it is now the accepted way of working and it is also accepted that it is a continuous improvement process itself.
One aspect that we found difficult was that Agile was a relatively new approach in business computing a decade ago and there was little experience in the private sector in delivering it at the scale that we needed. This led to some difficulties in marrying an overall architecture with ‘user-led’ computing. We placed some emphasis on business and technical architecture and ensured, after some stumbles, that we retained control of the design process which is fundamental to achieving an integrated result.
I think the approaches work well in practice and deliver business value at a regular cadence. As I’ve said, we have releases of new functionality on approximately a monthly basis and people can see steady improvements all the time.
When budgets are tight as they are now, is there a pressure on IT simply to deliver service more cheaply, or were there opportunities to innovate and make a more strategic contribution?
I’m not sure that the options in the question are fully mutually exclusive. There has always been pressures to deliver value for money and, as you might imagine, they have increased. The level of pressure has also resulted, though, in a number of fundamental reviews of expenditure; this has two positive effects – first, some recognition that longer-term benefits generally require investment and second, some questioning of conventional wisdom and practices. The latter means that we are more open to innovative solutions.
Public-sector IT sometimes gets a bad reputation. What’s your reaction to the international report from earlier this year, praising the DSP’s approach?
I’m very pleased with any positive publicity, though I was somewhat amused at how grudgingly some of that was expressed! I think that ICT in the public sector is at least as good as elsewhere and I would back some of my staff against the best in the private sector.
A few things distinguish public-sector ICT – it tends to be large scale, it tends to be public and it is still seen as something that can be managed and staffed by ‘generalists’. This means that any failures are relatively big, they are very public and they are generally due to failures of management or governance.
It worries me that, apart from the staff involved, there are two widely prevailing views of how to ‘solve’ ICT in the public sector. One is by deployment of private-sector software or services and the other is by better project management.
There is absolutely a place for such private-sector software and services in the public sector but, unless one believes that there is no difference in the sectors and that money is not a problem, they will not provide the specific support that many public-sector organisations need without significant and expensive customising.
As regards project management, project planning is an exercise in predicting and shaping the future and is inherently problematic. There comes a point where excessive application of the wrong sort of project management is counter-productive. I think a lot more emphasis needs to be placed on Agile practices and their management.
Do you think there are lessons here for others to follow?
I think that business systems are very similar to organisms. I think this means that an evolutionary approach is probably the best long-term strategy. Adoption of Agile, Lean and other people-oriented methods are fundamental to success. I think the methodologies need to be deployed with strong guiding principles, good governance structures and a continuous learning approach.
In many organisations, IT is often perceived as lacking the skills to understand the business and its direction. What can IT departments do to get to this level in more organisations?
I think that it is true that a lot of ICT people are quite ‘binary’ and literal in their approach to their business colleagues and this can come across as a ‘lacking’. I also believe it is true that information processing in its widest sense is a fundamental function and responsibility of any business area and that some business managers are lacking in their understanding of this. I think attention needs to be paid to both sides of the business/ICT interface.
On the one hand, ICT professionals who have a design or architectural role need to be involved where business policy, architecture or design is being formulated or responded to. On the other hand, business area professionals need to be able to articulate their direction and information needs.
I think that trying to implement approaches such as Domain-Driven Design with agile prototyping can go a long way to addressing some of the issues (ie, ICT needs to quickly deliver something that is obviously what the business wants and build credibility on the way. At the same time, business areas need to understand and engage in the process).
Did you or your team have to undergo any training or reskilling to take on this new approach?
I think that working successfully in the ICT field involves committing to ‘life-long learning’, so this was no different. We all require training and reskilling at times but also need to maintain professional expertise by reading, networking and following industry trends. We try to maintain a core emphasis on all these forms of research and then translating them into formal training and reskilling when we feel that they need to be ‘mainstreamed’. In this case, we had a mix of Agile, Object-oriented tools and techniques and BPM and we also had the advantage of working with Richard Pawson who helped guide us through initial implementation.
Are there other technologies that you think could really deliver value to the department – what’s planned for the future?
There are a lot of technologies that have a role to play in where we’re headed. The most obvious ones relate to mobile, BI/analytics and cloud and we are actively pursuing initiatives in all [of them].
On the Agile front, there are tools and techniques that we can investigate and adopt like Behaviour-Driven Development. All of these are strong candidates for adoption at this point and we will work to incorporate specific aspects into our practices and procedures as they prove their usefulness. We hope to advance on all fronts in the next few months (and years).
Can you tell us a little bit about your career to date: did you start on the technology side of things, or a business process role?
I started, a long time ago, in the Office of the Comptroller and Auditor General. I consider that the role of auditor is to listen, understand and critique systems and business processes – not a lot different than that of a systems analyst!
In the mid-’70s, I became intrigued by ICT as it was making an appearance in the areas that we were auditing and started pursuing a diploma in systems analysis. I picked up on ICT as a hobby at the time but then found that I was really interested in pursuing it professionally, as well. I joined Central Data Processing Services [CDPS] as a systems analyst and switched to a degree in computer science, which I finished in 1979.
CDPS broke up and I went to the Department of Social Welfare in 1984, where I’ve been since. I spent some time outside the IS Division when I took over the department’s central records in the Nineties but returned as director of IT in the early 2000s and have retained the CIO responsibility since.
How much of a bearing do you think your background has on how you approach the role now?
I’d like to think that it all helped! I think the early years of auditing and analysis helped me develop objectivity and a focus on overall objectives. My hobby and degree gave me a theoretical and practical understanding of the ‘harder’ side of IT and my time with business responsibility in the department helped me to understand some of the business and management issues.
Buy your tickets now!