If you were to start from scratch on a blank canvas, would you design your application landscape as it is now? Or would you do something different? Think different and make it better?
The modern application landscape is made up of a blend of applications running on-premise (or on dedicated infrastructure in a data centre) and in the Cloud, with business processes running across them which are mostly ambivalent to where the applications are hosted. Such applications come as SaaS or Cloud ready out-of-the-box, modularised and wired with a suite of APIs to integrate across a wider ecosystem.
For the most part, Start-ups have the luxury of beginning from a blank canvas. They utilise SaaS applications as much as possible and build their core Intellectual Property applications in the Cloud, leveraging IaaS and the PaaS components of the Cloud to simplify and vastly speed up product development, and reduce time spent installing, maintaining and upgrading systems. This provides Start-ups with the flexibility to quickly adapt and pivot their products and services to evolving customer needs whilst keeping operational costs to a minimum (paying for what they use).
It doesn’t need to be different for well established businesses. We live in a world where it is not the big that eats the small, but increasingly the fast that come out on top. It is imperative that all businesses, no matter what their size or years in operation, address the obstacles that block them from having the ability to continually innovate and adapt in response to change.
More often than not, the obstacles stem from two things; application baggage and resistance to change.
Over the years, the business may have accumulated a lot of applications, many of which roughly do the same thing, some are heavily customised with lots of moving parts and are embedded into the system which works across the business making it problematic to change. Process and applications are interlinked. Mergers and acquisitions will usually bring lots more duplicate applications into the mix. There is an application baggage problem that is inhibiting change.
The default position tends to be to just leave it as it is, don’t rock the boat, with the focus on managing and controlling the complex application estate over realigning the estate for improvement and future growth. The thing is that this position, hoarding technical debt, gets worse every year.
You do need to rock the boat.
In order to improve your business agility and bring new products to market faster that are continually adapting to deliver better customer service whilst also reducing operating costs, you’ve got to look towards a Hybrid IT model upon which to modernise your application and process estate.
UST’s overall approach to modernising the application landscape focuses on the wider transitional or transformational journey which applications will need to take so as to drive the greatest value from the application and hybrid IT estate. We often find that the business case highlights the opportunity to significantly self-fund much of the modernisation programme as each iterative release of the modernisation roadmap returns value and cost reductions back to business.
This journey is rarely a one-step move, but rather involving different combinations of routes for each application or application group. For example, one application might be refreshed and rehosted into a new data centre as a temporary measure, then replaced and decommissioned once a new system is ready. Another application might transition to the cloud, firstly getting re-platformed to leverage specific PaaS components and secondly rearchitected to truly leverage the suite of services offered by the cloud provider. This is illustrated in the figure below.
Each of the different steps can be summarised as follows:
l Rehost – Lift and Shift application to a different Data Centre or Cloud l Refresh – Upgrade underlying technology to latest versions for better maintenance, support and security l Replace – Replace application with COTS / SaaS product or Cloud Service l Retire – Decommission app following rationalisation and consolidation l Outsource – Outsource legacy or sunset applications through to decommissioning l Service Enable – Wrapper around legacy application to open it up as a service l Re-Platform – Switch out some components (minimal code changes) to different technology l Conversion – Convert to a target technology stack l Refactor – Code changes to improve design and performance l Optimise for cloud – Reengineer so as optimised for Cloud and leverage Cloud PaaS
There is often a temptation to simply move suitable applications first into the Cloud as part of a Hybrid IT strategy, and then later think about how they could be modernised to leverage the benefits of the cloud. However, that “later” sometimes never comes.
To paraphrase a quote attributed to Einstein, “If I had an hour to solve a problem, I’d spend 55 minutes thinking about the problem and 5 minutes thinking about solutions”.
Our experience within UST is that it is better to first build out a detailed understanding of the applications and how they support the organisation’s lines of business and associated business processes, noting any challenges and problems with how it currently works. This in turn enables a sequenced roadmap and timeline for application modernisation to be assembled in a way that will drive the greatest value for the organisation, with a business case that addresses outcome factors such as improved productivity efficiencies and performance, enabling of innovation and business agility to deliver better customer experience whilst also driving down costs. This strategy would also look at process modernisation, and how existing processes based on legacy constructs could also be greatly improved and automated in combination with modernising the underlying applications.
Once the high-level thinking and design has been done up-front, the implementation is usually not “big bang” but rather one of continuous application and process portfolio improvement. We look at how the outcome value from each iteration can be leveraged to contribute to the funding for the next iteration.
Where the use case makes sense, low-code development can also be a great way to modernise some applications. We see low-code development rapidly accelerating over the coming years, in some ways acting like a “juggernaut” that will hit traditional software development.
In conclusion, Hybrid IT makes business sense, but don’t just do it as a “tick the box” exercise, do it as a means to modernise your application and process landscape. Revert back to your blank canvas design and move towards it.