New approaches in the project management of large-scale IT projects
Every IT executive struggles with the implementation of large-scale IT projects. Studies consistently show that two out of three large programmes regularly fail to meet the original budget, miss the schedule and fall short of user goals and expectations after a long period of implementation. Often by a multiple.
These lapses are particularly detrimental to companies, as large programmes are usually of critical importance. Whether it is a global consolidation of ERP systems, the system-wide harmonisation of operational processes, or the merging of financial data across locations.
A number of new methodologies and tools that have emerged in recent years can significantly improve the implementation of large programmes. Combined with disciplined and experienced project management, the success rates of complex large-scale technology projects can be significantly increased.
We’ll show you how.
We show you below how programme management can be supported with the help of modern and partly digital applications.
Using agile methods
Even companies that are committed to agile development often resist using agile methods for large programmes. This is not about black and white thinking, either agile or not, but rather using the advantageous components of agile management for themselves. From our experience, these are:
- Clear product responsibility
- Prioritised product backlog
- Roadmaps based on sprints
- Small cross-functional teams
- Iterative releases with timebox sprints
- Modular architecture
- Use of Objectives and Key Results (OKRs)
- Minimal Viable Products (MVP) as the first major milestone
.
.
.
An agile mindset also encourages project staff to work more closely together, to break down possible silo thinking, to react more quickly to change, and to keep testing and learning.
The use of agile methods in large-scale projects also brings some advantages. The conversion of programmes that have already been running for some time to agile development also pays off – despite the late timing of the conversion:
- Faster to visible results
- Lower frustration among users
- Fewer errors in the system
- Better collaboration between different providers
- Inclusion of smaller development companies instead of a large system integrator
- Increased productivity
- Higher delivery speeds
.
.
.
.
.
Using design thinking
.
Many large programmes meet functional requirements but not the true needs of users. To meet the users’ requirements, design thinking comes into play. This is a method of problem solving that helps deliver products and features that users want and need and are therefore more likely to use. Another benefit is savings, as teams only develop the features that are really needed.
For large programmes, design thinking starts with identifying user requirements, through a mix of quantitative survey and qualitative research. This gives a clear picture of how users want to use the product, what their experience has been so far and what needs have not been met so far. It would be a mistake in a long-running programme to see this as a one-off. Rather, it needs regular and intensive user involvement throughout the programme’s life, for example in the creation of prototypes and in user tests. The comparison with the traditional approach should make clear how design thinking can work.
In the traditional approach, the user would be involved in the design process.
In the traditional approach, one would start with the collection of requirements from R&D, production, sales and customer service. This collection is often done in isolation from each other and then recorded in a requirements and specifications document. This is usually so complex that a review by the technical department more or less follows.
In contrast, within the framework of design thinking principles, cross-functional workshops and interviews take place to collect current problems and requirements together. Today, the use of a “digital twin” (Digital Twins) can be helpful. This is a digital simulation of a process on the computer. This makes it easier for those involved to identify problems and necessary interactions. This leads to an early exchange of information between the departments concerned, users, suppliers, providers and developers. On the one hand, an understanding of the true requirements is created for all, and on the other hand, the basis is laid for close and successful further cooperation during the course of the programme. This approach also speeds up the release of functions and facilitates the prioritisation of requirements in the product backlog by the product owner.
Cloud-based services
.
Many executives tend to reduce the benefits of the cloud to efficiencies in infrastructure management. However, the possibilities inherent in cloud services can also be seen when applied to complex large-scale projects. The features offered by many cloud service providers give programme managers flexibility in deploying different environments. Cloud solutions make it possible to scale development environments within a short period of time, allowing prototyping or testing of novel solutions in a short period of time. The creation of project reports, which is usually resource- and time-intensive, can also be done much faster with the help of the cloud.
On the software development side, using a Platform-as-a-Service (PaaS) solution can significantly increase productivity and provide access to thousands of innovative services.
But it is not only during project realisation that the cloud can be used sensibly; even when determining the basic solution architecture, numerous cloud services now open up new options. Opting for a software-as-a-service (SaaS) solution, for example, can avoid the effort of custom development and lead to an equivalent solution that is also easier to maintain.
More flexible through microservices and modular components
.
Many companies are moving to more modular, flexible architectures that allow for the inclusion of microservices, for example. This move allows companies to rely on multiple vendors in a programme, reducing dependence on just one major implementation partner. This also solves one of the biggest problems in managing large programmes: with single-vendor solutions, it is almost impossible to exert constant cost pressure, as fixed-price contracts often include a significant risk premium and change orders are common. On the other hand, time-and-material contracts create incentives for providers to extend or expand programmes and thus increase their revenue streams.
Modular architectures, on the one hand, allow clients to collaborate with multiple providers as mentioned, and on the other hand, individual providers can be replaced as needed. If several development companies work together in the programme, they can either compete against each other in each project phase or each receive smaller work packages (microservices), e.g. front-end designs or development and testing services. As the programme progresses, the best-performing providers – who bring in consistently competent staff at a reasonable cost – can then win the contract for a larger portion of the work.
Conclusion
.
Large-scale IT projects continue to pose immense challenges to corporate executives. Their implementation has a significant impact on company operations. Not least, such programmes require massive budgets over several years, tie up a large number of employees in the company and are crucial for the careers of many executives involved in the project. Whether such programmes are successful certainly depends to a large extent on the leadership qualities of the people involved. In recent years, however, new approaches have been developed and technical possibilities have been added that project managers can access. Those who use these possibilities and apply them to their programme significantly increase the likelihood of success of complex projects. The best thing to do is to discuss with the project managers at Arentzen & Partner the extent to which the approaches presented could also benefit you and your project. Contact us!
Images used: