How to make the most of a composable enterprise architecture

Championing composability

Today’s business mantra is one of agility. In this context, the IT sector is coalescing around the concept of composable business. Accenture urges CIOs and CTOs to be the champions of composable tech by configuring and reconfiguring business-critical applications, while still ensuring interoperability.

But traditional approaches to enterprise systems have made composability challenging. According to Nick Jewell, technology evangelist at Incorta, enterprise resource planning (ERP) systems are, to some extent, the antithesis of this composable model.

“ERP systems are large, monolithic platforms that govern mission-critical business processes such as order management, transaction processing or running supply chain operations,” he says. “Getting data out of such complex platforms often involves significant IT investment in data architecture, such as extract, transform, load (ETL) tools to move, transform and deliver data in a usable form for analytics and data-driven decision-making and data warehouses to hold that data over time.”

With composable IT, the aim should be to integrate ERP data with other technology solutions so the business can benefit faster. According to a recent survey of 503 CIOs and 503 CTOs conducted by Censorwide for Rimini Street, 84% plan to make investments in composable ERP in 2023. Both CIOs (83%) and CTOs (85%) are enthusiastic about investing. The survey shows that IT leaders in manufacturing (93%) have the largest commitment to composable ERP, whereas utilities (23%) have the smallest.

Research from analyst Gartner shows that by 2024, 70% of large and mid-sized organisations will include composability in their approval process for new applications. For Gartner, this means using an enterprise IT architecture for business applications based on modular building blocks.

According to Gartner, one of the key differentiators of the composable enterprise experience is that application design and redesign are performed with the direct participation of business and technology professionals. This suggests business and IT people need to work in tightly integrated teams.

Finding flexibility

Damian Smith is CTO at Podium Analytics, a non-governmental organisation and charity founded by former McLaren Group chief Ron Dennis in 2019, which aims to reduce injury in sport.

Podium Analytics has been gathering data using an application that records injuries in school and club sports. In September 2022, it introduced a tool based on the Concussion Recognition Tool (CRT5), a Concussion in Sport Group protocol designed to help non-medically trained individuals identify suspected concussions. The application provides guidance for removing a player from play and seeking medical assistance. 

One benefit of a composable model for business is improved flexibility. This works at various levels. Podium Analytics uses the low-code platform from Outsystems to provide this flexibility. Even though it outsources software development, the platform enables Podium Analytics to retain ownership of the generated code.

By using a low-code platform, Podium Analytics gains business agility. It can introduce more developers or replace its existing outsourced software development provider much more easily. “Low code enables another developer to pick up somebody else’s work really quickly without all of the kind of the archaeology of trying to work out what’s in the code,” he says.

The essence of agility

An application is effectively used to solve a business problem based on the data it has access to. Explaining how an IT leader can apply this when building out a portfolio of enterprise software, Accenture’s McGuigan, says: “It’s less about comparing one enterprise software provider with another.” Instead, she sees CIOs looking at their enterprise as a whole need to consider having a strategy that gives them the ability to unlock data housed across hundreds of applications.

As an example, Smith says Podium Analytics may decide it needs to build a new application to enable a coach or teacher to record a particular type of sports injury. “We’ll build an app that enables them to do that,” he says. “They can then record these injuries and we have data coming in. At some stage in the future, there may be a better way to gather that data, but we don’t really mind.”

Such a strategy can be applied across the technology infrastructure businesses rely on. The theory behind composable business is that every component of the architecture can be swapped out if needed. Giving an example of how this may work, Smith says: “We use a CRM [customer relationship management] system, but we are not precious about which one we use. I don’t care because they’re all doing roughly the same thing.”

If, at some point, Podium Analytics decides to change its CRM provider, the main task will be migrating data over to the new system. Smith’s view of composable business is that it should be like a Lego brick model, allowing IT decision-makers to choose which module they need to build out their IT architecture. “I should be able to take one out and put a different one in without too much disruption to the business,” he says.

Adding agility

An agile software development methodology goes hand-in-hand with a composable business strategy, since it enables business leaders to go to market with new ideas quickly, test if they work and tweak them where necessary. Podium Analytics’ development methodology is based on a four-week sprint cycle.

Describing the software development process, Podium Analytics’ Smith says Outsystems uses flow diagrams, which allow the programmer to describe what is presented to the user, what happens to data that is keyed in, and what screens are then displayed.  Even though there is no formal documentation, he says, it’s easy to see what’s happening in a piece of code.

Cherry-picking components

With the advent of software as a service, Smith says there are plenty of opportunities to assess best-of-breed products that have the potential to work exceedingly well in certain business processes. “Why would I bother to do a massive ERP implementation when I can just take one of these products and plug them into our IT architecture so that all the components can work together beautifully at a fraction of the cost?”

This strategy allows IT leaders to take the best technology available either from established software providers or startups. But, says Smith, if you choose to take this approach, you need to be really conscious of your exit strategy. “If something better comes along, or we fall out with the existing provider, we just want to take that Lego brick out and put another one in its place,” he adds.

The question IT leaders need to consider is the ease with which they can export the data from their existing provider’s systems and whether this data is in a format the business can easily use. Smith says this policy is fundamental to Podium Analytics’ procurement process and is one of the reasons he selected Outsystems over other low-code platform providers.

In these uncertain times, Emmanuelle Hose, EMEA general manager at Rimini Street, says businesses do not want to be locked into just one way of doing things. They need, she says, the ability to be much more nimble, and this is changing their approach to digital transformation. “You need to have the flexibility to change very quickly due to economic challenges.”

To avoid lock-ins and make the most of existing and new data sources, IT leaders need an enterprise architecture that can support the business in a constantly changing environment and is able to react quickly to changes. From an IT perspective, this is the goal of a composable business software strategy. However, this agility comes at a cost – it adds complexity and IT departments will likely need to spend more time managing multiple supplier relationships.