Digital projects need to reduce dependency on specialist digital and programming skills, according to Jane Roberts.
Digital transformation is a lofty ideal. The rationale is to emancipate and empower citizens and civil servants not only by transferring paper processes into the digital medium but to improve the way services function altogether via that medium. That’s a big ask, as it requires multiple personnel to work together to design, construct and deliver. So it should come as no surprise that there have been numerous sticking points that have challenged projects over recent years.
Number one on that list has to be cost, with many projects going over budget and over deadline, which is why the Major Project Authority (MPA) was brought in, to try and keep projects, listed under the Major Project Portfolio (MPP), on track. A recent survey by the National Audit Office (NAO) suggests 35 percent of projects (37 out of 106) of the projects scheduled for delivery in the next five years in the MPP are likely to overrun their deadline. The reasons cited include “high turnover of leaders, a failure to attract vital talent, and skills shortages relating to digital and commercial expertise forcing projects to rely on contractors”, leading NAO head Amyas Morse, to conclude “… the overall picture of progress on project performance is opaque. More effort is needed if the success rate of project delivery is to improve.”
There’s little doubt that much of that obscurity is caused by a dependency on the knowledge and technical capabilities of the large systems integrators (the ‘oligopoly’); a dependency which some contractors have sought to perpetuate. Large SI’s can and do have staff with highly specialised skill sets, making their job function very narrow, which can increase the manpower needed on a project. Consequently, GDS has pushed for departments to utilise in-house resource. If the skills needed aren’t available internally, the project leader may well embark on a recruitment drive, such is the aversion to outsourcing, but this again ramps up cost. A multi-disciplined resource is needed and it’s for this reason that the SME is gaining in favour and will continue to do so as more departments go on to GOV.UK.
G-Cloud sought to connect the SME with these departments and foster a greater sense of competition to loosen the strangle hold of the large SI. But what can fail to come across is the value added. SME team members are typically able to turn their hand to several different tasks, which can be a real boon when it comes to keeping digital project costs and days to a minimum. One example of this is agile coaching. Rather than bring in a dedicated agile coach it makes far more sense to appoint a Project Manager with this skillset. Staff with a broad skillset can also help to ensure continuity, for instance if those responsible for project planning and strategy also have the skills required to test the implementation, they have a wider understanding of the overall design requirements. So a Test Analyst familiar with design requirements and analysis capture is able to take a far more holistic approach.
Of course, it’s not just a matter of manpower; there’s also the issue of technical literacy, or the ability to configure, integrate and implement solutions. One of the key issues thrown up by the NAO survey is the difficulty government faces with digital and commercial expertise. Put simply, the digital process itself becomes a barrier to the development of services because many of those on the team won’t have the necessary coding skills. A large part of the digital delivery process is focused upon developing iteratively which are then subject to iterative testing but the technical nature of the process can see this become the preserve of programmers. Non-technical team members can find this highly frustrating as their contributions are restricted to workshops and collaborative reviews, adding time and complexity to the project. A far better way is to encourage a collaborative working practice but in order for this to happen there needs to be greater transparency. Low code solutions, which enable team members with limited coding capabilities to view, influence and even take a ‘hands-on’ part in the process, are one way of promoting collaboration, and can ensure working prototypes are developed from day one.
Control is key
Ownership of the project is also key to keeping down costs and there are two ways to do this: self-configuration of services and usability testing. Self-configuration enables the government organisation to take control over the way services are adapted and deployed, enabling existing services to be easily modified or repurposed. The Charity Commission, for example, is currently developing reusable service prototypes and have a skills transfer plan in place which will enable the in-house team to own and activate digital services in the future without the need for external technical support. Autonomy has to be the way forward for both large and small departments.
Usability concerns how easy it is for that service to be accessed over multiple media. Interoperability and responsive design ensure a consistent user experience. If the solution needs to be reconfigured for multiple devices, costs can spiral due to the need for extensive testing across different platforms (mobile, tablet, PC) and browser versions. Dynamic solutions that automatically reconfigure and resize the display according to the device being used can dramatically cut project costs by removing the need for multi-platform and multi-browser testing.
Finally, there’s the problem of scope creep, as unexpected issues and delays cause projects to go over deadline. The mistake many departments make is to assume that the agile design process will accommodate adjustment at any stage of the process. This simply isn’t the case, for while agile does provide substantial flexibility and room for manoeuvre, changes need to be agreed before implementation, otherwise timescales and budgets can run wild.
Digital transformation has the power to create a more dynamic form of government. But there’s been a tendency to rely heavily upon project leaders, external suppliers or the process itself to provide steerage and direction. If one of those lynchpins is removed, the project falters. In order to improve success rates, digital projects need to begin to reduce dependency on specialist digital and programming skills. Budget overruns suggest this is not happening so there needs to be a radical rethink. Focusing just on transformation as the end goal is no longer sufficient; it has to also be the transfer of the knowledge needed to bring that transformation about.
Jane Roberts is strategy director at supplier Toplevel