Andy Scott, Client Services Director, Jumar Solutions Limited
The simple answer to that question is: that you have good reason to be. The phrase ‘too good to be true’ springs to mind when being told that a complex CA Gen transformation project could be carried out in a fraction of the time compared to traditional development methods.
The scepticism tends to manifest itself in two ways; suspicion over whether using automation is viable – and suspicion over whether it is genuinely possible.
In this blog, we’ll attempt to address those questions – but if you want the short answer it is, respectively, ‘probably’ and ‘yes’.
With any complex modernisation or transformation task (be it CA Gen or another technology) there is a tendency to think that your particular circumstances are so unique and complex, that no automation software could possibly produce the desired results. Over many years we’ve seen dozens of such environments, and can confidently say that there is no such thing as a ‘standard’ set of circumstances; we have, however, seen a diverse range of implementation styles and ‘standards’. Systems have typically evolved over time, they are often business critical and they may follow standards to greater or lesser degrees. Despite this, highly-tailored automation is a perfectly viable option for modernising and/or re-architecting legacy and ‘tightly-coupled’ systems. The best way to ascertain if this really is the case, would be to look at three basic elements included within the preparation / planning phase for the project.
- What does the system look like now?
- What do you want it to look like in the future?
- What steps are necessary to achieve the desired outcome?
Scenarios where an organisation can create a well-defined set of required transitions, based on a full understanding of the existing ‘as is’ configuration, best lend themselves to the application of tailored automation. But that’s not to say that more complex scenarios don’t. More on that, later.
Scale is also an important factor to consider here. Much of the investment in tailored automation is done upfront – and if there are, say, thousands of in-scope objects to be processed, then that upfront investment is effectively distributed across that large number of objects. The implementation of a risk mitigation strategy to approach the project in a logical phased manner (which is where the majority of our expertise lies) and the resulting economies of scale should help to mitigate any fears or suspicions over the viability of using automation in these circumstances. When projects are done on such a large scale, we are effectively industrialising the process and supporting the required transitions via automation, which provides greater predictability and high quality resulting in cost effectiveness for the client as the time and cost to process each in-scope object reduces dramatically.
We’ve touched on this before but, vital to understanding the feasibility of using automation, is to understand what you currently have. In our experience, only a very small number of organisations know EXACTLY how their system is set up (and how it has evolved over the years following original development), and the majority don’t. (It’s nothing to be embarrassed about – read our previous blog for the reasons why). Even on occasions where an organisation DOES know exactly what they have, it is still important to carry out a comprehensive model analysis exercise to confirm / validate the ‘as is’ situation, hence ensuring that the automation, when tailored and executed, absolutely delivers the desired result.
This understanding of the target ‘to be’ scenario needs to be thoroughly defined, in order to determine the transition steps that are candidates to be automated. It’s not unusual for organisations to have defined the target architecture and supporting technologies, but without specific knowledge of the capabilities of automation software, there is a danger that opportunities are overlooked or dismissed as they are not considered viable when implemented using traditional techniques. Our approach encourages clients to stop and take a methodical approach to the process. Only now can the viability of automation start to be considered in earnest and a cost-benefit justification be prepared. However, as a general rule, in our experience, in the vast majority of cases where the scope is large – it is viable.
Now, we look at the second major area of scepticism. Is it really automated, or simply outsourced to a low-cost high-capacity development shop perhaps located on a different continent?
We can’t speak for other organisations that offer automation solutions, but at Jumar, our automation software does exactly what it says on the tin (read more about our automation tools and Project Phoenix), and we can prove it with strong customer references from complex “real world” CA Gen projects. All automation actions are recorded and documented in a comprehensive execution log report – where the speed of completion of each action is blatantly so fast and the number of objects processed so large that it’s simply not possible that it is being done manually. It is only when looking at such a report that the magnitude of the savings made by using automated methods really becomes clear. Imagine how long it would take (and how much it would cost) to do it manually, or to introduce a change at a late stage in the process. Not to mention the very real risk of introducing manual errors and inconsistencies. Highly tailored automation is fast, consistent, predictable and reliable – operating across the entire implementation as specified it can subsequently be adapted and re-executed as required.
Degrees of automation
One final thing before we draw to a close on the subject of automation – it may be that ‘degrees of automation’ need to be considered when defining the transformation roadmap and the approach to be adopted. There have been occasions where we have been faced with a particularly complex situation where a degree of pragmatism comes into play. For example, when supporting a customer with a migration from a very tightly coupled implementation to a new functionally isolated “n-tier” architecture. Upon further investigation, separating the application into its constituent tiers may show that, in practice, even with automation software, it may be viable (and cost-effective) to automate most – rather than all – of the task in hand. In a recent case study, the tailored automation software effortlessly dealt with 80% of the required transformations, but it was decided that the remaining 20% would be tackled manually. The net effect was still a considerable time and cost saving over 100% manual work. Where automation is not viable (and cost-effective), there’s no way we would recommend it.
If you have a CA Gen – or a related legacy technology – transformation project (planned or underway), I’d be more than happy to discuss the benefits that automation could bring to your organisation.
Just drop me a line and we can talk some more.
I’m more than happy to attempt to lay any suspicions to rest once and for all!