Main content

Put an API-centric strategy at the heart of migrating your core platform

Matthias Steiner Profile picture for user matthias_steiner October 26, 2020
Upgrading a core ERP system is typically a multi-year project - but an API-centric strategy can advance transformation goals along the way, writes Neptune Software's Matthias Steiner

Software Computer program upgrade business technology internet concept © Alexander Supertramp - Shutterstock
(© Alexander Supertramp - Shutterstock)

Businesses around the world are facing the huge challenge of updating their core systems to new intelligent technology platforms that are fit for a modern, digitally enabled enterprise. The paradox that IT leaders have to reconcile is satisfying the demand from business colleagues to see quick wins, at the same time as planning and executing a multi-year technology migration.

The stakes are high as deadlines loom. Customers demand more timely, responsive and engaged interactions. Digitally connected competitors are emerging to vie for their business, putting pressure on enterprises to complete their own digital transformation. At the same time, the core systems established enterprises have relied on to power their business for the past decade and more are approaching a time when vendors will no longer support them for maintenance and new functionality.

As an example, a recent estimate by diginomica found that some 32,000 SAP customers worldwide run on core ERP software that is set to lose mainstream support in 2027 as the vendor switches its attention to the latest version running on its in-memory HANA database technology. SAP user groups are advising their members to take this as a cue to move forward with their modernization plans. Andreas Oczko, Board Member for Operations/Service & Support at German SAP customer group DSAG says:

The maintenance commitments for S/4HANA through at least 2040 and for Business Suite 7 through the end of 2030 are not a carte blanche for companies to keep waiting. Instead, it is a sign they should put their recent reluctance to one side and start tackling digital transformation. 

A modular transformation plan

The problem for IT leaders is that modern IT transformation projects do not happen with a big bang. A  multi-year core systems upgrade that will start to deliver results 3-4 years from now isn't going to get signed off, especially with all the current economic uncertainty. Indeed, in its latest earnings call, SAP's Adaire Fox-Martin confirmed that large scale projects that were in planning are largely off the table and that customers are looking for smaller projects that will deliver quick wins. In turn, this means that project leads must understand how to convince key stakeholders and ensure executive sponsors buy-in. The best way to get trust by executive management (and peace of mind for yourself!) is to constantly deliver tangible and innovative digital solutions along the way.

Typically, a project to modernize a core system is part of a broader digital transformation strategy in which the organization's anticipated IT landscape is changing in various ways to address line-of-business requirements. This means there will be multiple facets to the project — not just upgrading the core system, but also catering for functionality gaps in the new system compared to its predecessor, and a need for differentiation through custom development. There will likely be other systems such as CRM and customer engagement, supply chain and spend management, ecommerce, revenue management and analytics. 

In this scenario, the old proposition of a single core ERP system with its own proprietary technology stack is outdated. The modern approach being adopted by vendors and their customers is a loosely-coupled, modular suite of applications. The core may still come from your trusted vendor, but there will be many satellites and planets. The mechanism that you use to orchestrate this modular system is the key to being able to deliver quick results at the same time as progressing the longer-term core upgrade.

An API-centric strategy

An API-centric approach allows you to add a mediation layer that can work with your current and future core systems as well as integrating satellite applications and providing a framework for building new functionality. Developers used to working with your core systems can model and define the processes and business logic and expose it in a secure, reliable and automated fashion as APIs. The API-first approach allows developers with other backgrounds to build new solutions, or extensions to existing ones, that access these functions. That makes an API-centric approach attractive to developers who can use the best tools for the most appropriate use cases rather than being siloed into a 'one-size (allegedly) fits all,' which we know from bitter experience is not the case in the real world. 

At the same time, these APIs act as sort of a façade. You can develop apps now — in parallel to the ongoing core system upgrade project — and then renovate the underlying implementation once the upgrade has been completed. This makes it possible to deliver modernization and the quick wins business leaders demand while continuing to make progress on the larger migration.

This approach becomes stronger when coupled with a development platform that can buiild applications the modern way in double-quick time. These modern platforms deliver web and mobile apps without detailed knowledge of coding norms like Javascript, HTML5 or mobile frameworks. Now your ERP developers can develop new web and mobile apps from within the core system without needing an additional platform, tools or skillset. Used in an ‘app-factory' approach this becomes a strategic in-house custom development platform for the entire workforce, supporting the development of anything from HR employee self-service (ESS) to manufacturing, plant maintenance, warehouse management and so on. Too good to be true? Not at all.

It really works

We are seeing customers from many different industries unlocking the value they know exists within their systems but with which they have previously struggled to access let alone realize. Check out Den Howlett's story about Eastman Chemical and its improvements to 'wrench time:'

It works, it's stable and we've reduced training time from something like 40 hours I mentioned earlier to about two hours. That's an important productivity gain for us. But equally, we've made the user experience one where the user wants to use the system so in the end, we get better and more reliable data too.

Or look at how Hitachi Rail transformed its SAP ECC 6.0 implementation from a system that nobody wants to use, to one in demand from the business:

For example, before developing the stock take and inventory app, users were going to bins, doing manual counts on paper, they were then going back to the terminal and then entering that information in the SAP back-end and posting it there. 

Now, they literally pick up a hand scanner, they go to the bin, they scan through the materials, and the information is integrated straight away. It’s a faster process, we’ve removed the double entry, but also the transit time across the warehouse. We’ve seen a big increase in the amount of counting they can actually do.

As you might imagine, we have plenty more stories like these. 

In conclusion, we believe this recipe — taking the best of the old and the best of the new — allows you to develop apps in a future-proof way that leverages the existing skillset of your workforce, while running in parallel to an ongoing core systems upgrade. Read our white paper for an example of how this works in the context of upgrading to SAP S/4 HANA.

This story was updated to provide some examples of what customers are doing and to reflect what diginomica heard on SAP's latest earnings call and in checkpoints with other analysts. 

A grey colored placeholder image