How integration can make-or-brake your digitalisation effort?

As organisations are accelerating their efforts to digitalise their products and services, adopting more SaaS capabilities, transforming their applications to cloud-native services, analytics are being embedded into business's processes and customer interactions, both application integration and data integration strategy can play a critical role in the success of digital transformation initiatives.

Most of the traditional enterprises have hundreds if not thousands of applications, with a mix of home-grown and COTS solutions and with increasing number of SaaS solutions being implemented, integration landscape is becoming more complex than ever. It's critically important to have a well-defined and governed integration strategy that enables digitalisation but also addresses legacy integration challenge and defines a clear pathway towards becoming a flexible, adaptive and modern enterprise.

There are few key ingredients of any integration strategy:

Enablement of digitalised customer experience: Providing a delightful, personalised customer experience across any customer touchpoint is one of the key attributes of any digital enterprise. Service enabling your legacy applications by exposing APIs is a good starting point but care must be given to ensure that those APIs are reusable across channel, they are secured, performant and scalable. Managing those APIs in terms of their lifecycle, separating experience APIs from process APIs to ensure that customer experience can be changed without impacting your backend applications. API management capabilities like security, version management, service level management are important here.

Enablement of legacy applications : Most of the traditional enterprises have multiple integration platforms including enterprise message and service buses. Using these platforms to expose APIs seems like a natural choice but there are a number of pitfalls that one should avoid. In majority of the cases, these platforms suffer from huge complexities that have been added over a number of years. Business logic has been wrongly implemented on these platforms, duplicate services are implanted and generally they have become a bottleneck. As underlying applications themselves are transforming into cloud-native or being replaced by SaaS solutions, role of ESBs should be minimised and some cases fully avoided. One of the common mistakes, I've witnessed is that these cloud-native servicer or SaaS solutions are implemented with very tight legacy coupling. This has been done either by integrating new services via ESB or directly integrating with legacy. This approach will inevitably bring new digital services down as legacy will continue to suffer from time to market, performant and scalability. This is where well-defined data integration approach along with implementation of service-mesh architecture can help.

Data & Application integration, two sides of the same coin: Creating new cloud-native services or implementing SaaS product, will invariably require access to legacy at least during the transition period. One way of minimising this coupling with legacy is to keep the legacy data nearer to these services by implementing 'Data integration hub' which will ensure that all the changes in the legacy data sources are synchronised with the new services. This approach is particularly important where your new cloud-native services are acting as 'Source of Truth'. As cloud-native services are characterised by 'eventual consistency', ability to integration services via enterprise message bus and implementation of event-driven architecture is an absolute must. It's really important that data and application integration strategy go hand in hand. Real time data integration is no longer nice to have but a must for any digital enterprise.

Rise of Cloud/IoT : Traditional large enterprises are finding themselves with a mix of large on-premise applications, ever increasing SaaS adoptions and their own applications moving into cloud. Without a proper integration strategy, these enterprises can end up with significantly complex integration architecture. Multi-cloud services are becoming norm now and how we architect, design, implement and most-importantly operate these end-to-end services can become challenging. Having the right architecture, governance and technology (message bus, API management, service-mesh, light-weight integration frameworks, data virtualisation, CDC etc.) is critical. Most of the enterprises may have these technologies but without clear integration strategy, architecture and governance, it will provide only partial solution to the problem.

Collaboration across products and ecosystems : Creating new business models and revenue streams by patterning with other ecosystems is a hallmark of successful digital enterprise. This requires ability to integrate with partners and ecosystems, on-boarding their services, becoming an aggregator of those services, managing processes that span across multiple enterprises is all underpinned by a good integration strategy and capabilities.

Governance : Last but not least, one of the critical elements of any integration strategy is have an effective governance that ensures that architecture patterns ( that guide the design of services) are defined, communicated and implemented, service reuse is actively promoted across products and services ( particularly important as enterprise move towards product-centric model), automation across service life-cycle is implemented, security and service-level management policies are applied across all service exposures.

Integration was and will always remain critical element of any technology strategy but in this hybrid world of infrastructure, application, data and as boundaries between different industries are becoming blurred, it will make or break your wider digitalisation ambition.