Agile takes over big bang on the road to transformation

With organizations’ increased adoption of agile, transformation programs are moving away from the big bang waterfall model.

Ilampooranan Padmanabhan Apr 04th 2017
Big bang migration used to be the preferred approach for most of the core banking transformation programs in the past. The banks and financial institutions would adopt a waterfall model to migrate from their legacy system to a newer system. A typical implementation cycle would span a period of two to four years depending on the complexity of the IT systems involved. But all tangible benefits would manifest only at the end of the implementation cycle, provided the implementation was successful. And this was how the core banking transformation space worked for decades. 
However, with organizations’ increased adoption of agile, transformation programs are fast moving away from the big bang waterfall approach. And there are numerous factors which have influenced this change to a more agile delivery model.
  • Need for greater IT collaboration: Organizations world over have realized the need for self-organized teams with better collaboration, ownership and accountability of the deliverables. The aim is to develop better collaboration among the IT, business, and development teams.
     
  • Complexity of IT systems: Given the complexity of the IT infrastructure, data, multiple interfacing applications and short releases, module wise migration is simpler to handle than the big bang model. The lessons learnt can be immediately implemented in the subsequent releases.
     
  • Tangible benefits: Shorter and iterative cycles delivered by smaller groups result in tangible benefits to the business, faster than the programs that run for several years without any tangible benefits. There have been numerous instances when companies have embarked on multi-year transformation programs that ended with disappointing results, and have consequently seen the wisdom in adopting a model that enables constant appraisal and tracking.
     
  • Production changes: In a waterfall approach, only regulatory changes are allowed into production. All requirements need to be frozen well ahead of the implementation.  This makes it impossible to incorporate new changes which might occur from any changes in the technology. It results in huge losses and gives birth to an end-product that doesn’t cater to the needs of the business. 
How agile impacts quality assurance
Whereas conventional transformation laid emphasis on domain and application knowledge, in the agile model, automation and technology-led testing are at the forefront. The emphasis on minimal time-to-market promised by agile has precipitated this. The sequential model of software development which defines the waterfall model has been done away with in agile. This primarily addresses the banking industry’s need for quick-to-release applications through a number of derivatives such as SCRUM, SAFE, TDD, and BDD.
Agile involves testing at the API and web services level and is hence, a clear departure from front-end testing of core and interfaces that entails heavy dependency on application delivery. 
In such a set-up, automation is mandatory right from the start of the testing lifecycle. Continuous delivery requires continuous integration, and so, web services are tested and automated right at the start, so that the same can be used in subsequent runs. Performance testing is also not relegated to the very end of the implementation, as every release requires performance testing.
Finally, mobile and internet banking are no longer mere interfaces where the original application is extended, they have become services in their own right.
Challenges in the adoption of an agile 
The 2017 World Quality report states that, “involving QA teams in the inception and sprint planning phases is still a challenge faced by organizations.” Especially, with the inherent demand for speed and quick releases, it is impossible to test 100 percent of the releases. The key is to test the right things at the right time. 
  • With the adoption of cloud services and mobile applications, today’s digital enterprise is more vulnerable, and thereby creates a greater need for security testing.
  • Data migration in a transformation program requires a different approach to align with agile delivery.
Key takeaways for organizations
  • Organizations should focus on knowledge sharing amongst projects and departments. Cross pollination is critical.
  • Businesses should be quick in identifying emerging trends and constantly up-skill employees with the requisite skill sets.
  • Focus on asset creation in the form of automation tools with reusability to save effort, time and cost.
  • Invest in research and centres of excellence in specific technology areas to enable faster adaptability to agile mode. 
The author is Senior Manager, Maveric Systems
Disclaimer: This article is published as part of the IDG Contributor Network. The views expressed in this article are solely those of the contributing authors and not of IDG Media and its editor(s).