Growing competition from specialist providers continues to disrupt the traditional banking model.
These firms are using new technology and an agile workforce to build niche products without the overheads.
More established firms with a broader offering must update legacy technology to compete. In addition, the growing burden of regulatory compliance and burgeoning data volumes add to cost pressures.
All this means banks are rethinking how they streamline costs. One area where they are making savings is in the testing of core banking platforms. Testing is essential when implementing, upgrading or enhancing a platform. The problem was that it typically involved many labour intensive and costly elements – but that is changing.
There are three ways to cut testing costs: automation, offshoring and outsourcing.
The emergence of testing software designed for individual core banking platforms has opened up the possibility of greater automation. The increasing complexity and size of systems means automation is more than justified; while there is still a small amount of upfront configuration, banks can achieve a greater return on investment.
Crucially, these automated testing tools come with a library of predefined, ready-to-use test cases, which can be used and reused across multiple, similar testing environments, avoiding duplication of resources.
Offshoring is a known option for reducing the cost of labour, but data security is a major concern. For adequate testing, banks need to use realistic data sets. They have to test as many scenarios as possible, which means using data to accurately represent a maximum number of situations.
Real data is the most reliable but banks are nervous when sensitive customer details move outside their own walls. In some cases, regulations restrict what they do with data and that means banks cannot send it outside their national jurisdictions.
This is an obvious problem for offshoring but there are two ways banks are overcoming the hurdle: synthetic and anonymised data. Synthetic data is entirely made up but as realistic as possible, while anonymised data changes the sensitive details in the original information. Both approaches combine realistic data with lower risks of confidential information leaking.