We recently finished a discovery to help us learn more about some of the biggest legacy problems across government.
Although we currently offer some advice on how to move away from legacy systems, user research from the Service Manual and Technology Code of Practice shows that organisations want more comprehensive information.
Each government organisation has a unique infrastructure and set of services, which in turn means they have unique legacy issues. We aimed to discover how we could improve government guidance for legacy migration, and the best format for these improvements.
What we wanted to find out
In our discovery, we wanted to get a better understanding of what people identify as legacy technology, and the problems it presents. We also wanted to find out what support GDS could give the cross-government community. At this point, we didn’t focus on finding a specific solution or exploring a specific content strategy, as that would limit the field of our research.
Our user research
Our research consisted of a mix of face-to-face interviews and workshops. We spoke to technologists from government organisations, big and small, central and local to make sure we got a wide a range of experiences and opinions. Organisations we spoke to included GDS, Network Rail, Home Office, NHS Digital and the London Borough of Hackney.
So we could get a more accurate understanding of the variety of business and technical issues faced by organisations, we made sure participants came from diverse roles.
Here are 3 things we found out during the discovery:
1. Defining legacy technology is tricky
By the end of the discovery, it was clear that we could not group legacy technology under one definition. Our research found that when people talk about legacy, they can mean any of the following:
- infrastructure
- systems
- on-premise hardware
- business and IT processes
- old digital services
There were some commons reasons why the participants considered technology as legacy. This was when the technology was:
- old and at the end of life
- no longer receiving support from the supplier
- impossible to update
- presenting unsolved and unsolvable problems
- inherited, with inadequate documentation for current users
- unable to meet current standards
- no longer the most efficient option, in terms of cost and technically
2. Migrating from legacy is hard and data is the biggest barrier
While migration is technically hard, teams know what they need to do. And, they often have established strategies to complete a migration. It’s usually non-technical factors that stop a migration project. We split these factors into 8 categories:
- Business and culture - the current processes and culture of the organisation.
- Commercial - the commercial decisions and spending models used by the organisation.
- Data - the organisation holds and how it uses it.
- Finance - the budget for IT departments and cost of technology and resources
- Guidance and policy - work already being done by the organisation that serves as a source of information for other projects, and the policies that organisations have to follow
- Resources and responsibility - the people, budget, equipment and corporate knowledge that organisations have available to them. Also, who owns the decision-making process, who owns the technology, who is responsible for the different aspects of running the organisation technology
- Suppliers - who provide technology and support for that technology
- Security - how to secure technology when migrating especially when an organisation is inexperienced using cloud-based technology
Of all these issues, participants consistently raised data as the biggest problem. Migrating an organisation’s data is complicated and the amount of work is often underestimated.
Organisations frequently find there is:
- a fear of data loss and data ownership
- concern about where to store data
- data in an old system and in a poor state
- a lack of corporate knowledge about the systems holding the data and how the data was encoded
- data duplication and difficulty in discovering the authoritative source of data
- difficulty in finding all of the services that rely on the data
- a lack of trust of new data coming and ‘polluting’ the legacy data
- difficulty in keeping all the different data sources synchronised
3. How organisations are dealing with legacy technology
Just as there are many different types of legacy technology, there are also many ways of migrating from it. And, there’s no one-size-fits-all solution. Organisations usually look at a combination of strategies depending on their size and the complexity of their infrastructure.
Most organisations have an overarching IT strategy in line with their business strategy but some are taking an alternate approach to legacy. A common tactic is to migrate the business away from legacy in small parts, rather than all at once.
As one participant said:
“I think we have to accept that there's going to be legacy stuff out there, and there's going to be unsupported systems. So it would be better to accept that we’ve got that and come up with strategies for how we're going to manage that.”
What we’ll be doing next
GDS can help organisations in several different ways based on the feedback we’ve received. On the basis of this initial discovery, we believe that we need to do further research on the following topics:
- a high-level strategy for dealing with legacy
- identifying suitable case studies
- identifying legacy patterns and anti-patterns
- creating a community to share information and help each other with legacy technology
To make sure you stay up to date with all the latest developments, you can sign up to alerts from the GDS Technology blog.
If this sounds like a good place to work, take a look at Working for GDS - we're usually in search of talented people to come and join the team.
