Published
Author
AI is rapidly changing the conditions for today's ERP systems. At the same time, many organisations are stuck with older systems that complicate the transition to the cloud, modern data management and new AI features. The question is not only whether to replace the ERP system, but also which strategy provides the best business benefits in the long term.
The development of artificial intelligence has changed the conditions for how companies and organisations can develop, streamline and manage their operations. AI is no longer a future experiment alongside the regular system environment, but is gradually becoming an integrated part of business processes, decision support and operational workflows. In this shift, the role of the ERP system once again becomes a strategic issue – not just a technical platform for transactions, but an enabler for business development.
At the same time, many organisations find themselves in a more complex reality. The existing ERP system may be older, highly customised and locally installed. It may still support the organisation's fundamental needs, but at the same time limit the possibilities of using modern cloud services, data flows and AI-based components. This creates a central management question: which path is right when technological development accelerates, but the organisation’s system landscape is not necessarily ready to keep up?
It is tempting to view cloud and AI as arguments for a rapid technology shift. The reasoning may seem logical: if the old ERP system hinders access to AI, the organisation should change systems and move to the cloud as quickly as possible. However, such an approach risks oversimplifying the issue. A system change is rarely an isolated IT activity. It affects processes, roles, data, governance, integrations, and the organisation's ability to change.
Modernising older, locally installed ERP systems therefore requires a structured strategy rather than a one-sided focus on the migration itself. The organisation first needs to understand the business value sought, which processes can be changed, what data quality exists, and the organisation's ability to realise change.
For organisations with older ERP systems, several strategic options emerge in practice, of which the following are some simplified scenarios:
Procure and replace the ERP system
To accelerate a system change and quickly establish a cloud-based architecture that creates better technical conditions for AI.
Wait and await increased AI maturity among the ERP vendors
To wait a few years, with the aim of allowing both AI technology and vendors' own solutions to mature.
Upgrade to the latest version – prepare for the cloud
To upgrade to the latest version of the existing ERP system to close the technical gap and reduce the future step towards the cloud.
Add external apps as a complement to the ERP system
To complement the current system landscape with niche and modern add-on applications that can operate in a cloud-based context and also be prepared for AI.
None of these simplified options are generally right or wrong for the individual customer. The choice depends on the organisation's current situation, risk tolerance, financial capacity, technical debt, change management capability, and strategic ambition. The crucial question is not which system to choose, but which development path (read technical strategies) best supports the business's long-term goals.
One key insight is that a cloud-based ERP system does not automatically create improvements in the business. The cloud can be a necessary enabler, but it is not in itself a guarantee of more efficient processes, better decisions or higher productivity. Similarly, AI is not a universal solution that can be overlaid on an existing environment and immediately create value.
AI in modern ERP systems now includes functions such as forecasts, anomaly detection, automated posting, decision support and increasingly agent-based workflows. But the benefit of these capabilities depends on whether the organisation’s data is accessible, understandable and sufficiently quality-assured. It also depends on whether processes are mature enough to change and whether users trust the decisions or recommendations generated by the AI components.
The cloud can be a necessary enabler, but it is not in itself a guarantee for more efficient processes, better decisions or higher productivity.
Waiting can be rational if the organisation assesses that the AI functionality is not yet mature or relevant enough for its own operations. Vendors' built-in AI components are developing rapidly, and there may be reasons to wait for better standard features, clearer pricing models, and more proven applications.
However, waiting also carries a risk. Older local platforms can create increasing technical debt, complicate integrations, and make it harder to recruit or retain relevant expertise. At the same time, changing an ERP system often takes a long time and can stifle many other important business initiatives. If the lead time for an implementation is one to two years, every postponed decision pushes the timing for real modernisation even further forward.
A recurring challenge with AI in older system environments is not the AI model itself, but access to correct, consistent and useful data as well as the ability to feed insights back into operational workflows. This suggests that the issue of AI and ERP systems cannot be reduced to the choice of vendor or technology platform. It must be treated as an architectural and business question.
For many organisations, the most appropriate route may be not to choose between a complete system replacement and passive waiting. A third perspective is to gradually modernise the system landscape by supplementing existing ERP systems with new applications, integration layers, data platforms and AI-ready components. Contemporary development tools make it possible to relatively quickly create applications that improve user experience, automate subprocesses or open up data without the entire core system needing to be replaced immediately.
At the same time, such a strategy must be managed with discipline. More additional applications can also create increased complexity if they are not based on a clear integration and data model. The organisation therefore needs to define which parts of the ERP system should be a stable core, which capabilities can be placed in supplementary solutions and which areas should be modernised or replaced in the long term.
Organisations already using large ERP systems such as SAP, Oracle, IFS, Infor, and D365 face an additional dimension to their challenges. Often, the company has invested so heavily in the ERP system and its platform that anything other than continuing forward with the same ERP system can feel difficult to justify.
And it is precisely these vendors who push very hard to get their customers to adopt their new features and services that include AI – even if the customer's core installation is old and perhaps does not at all reflect how the customer's business operates today. Then the question becomes whether it really helps to put tape on something old or if it is time to rip off the plaster and consider a new path forward.
Organisations facing a crossroads need to analyse both the business needs and the future role of the ERP system. Those who want to choose the right ERP system need to evaluate processes, data, integration needs, AI potential, and capacity for change.
Management should therefore ask a number of interconnected questions. Which business processes have the greatest potential to be improved with AI? What data is required to realise this potential, and where is it currently located? What limitations does the current technical platform create? What costs and risks follow from upgrading, replacing, or supplementing? And perhaps most importantly: what organisational capability exists to drive the change from idea to actual benefit?
A well-founded strategy should include several perspectives: a clear vision for business development, a realistic assessment of the current system landscape, a prioritisation of AI-relevant use cases, a roadmap for cloud and data modernisation, and a governance model that ensures that technology choices are anchored in business needs.
AI and cloud technology are changing the conditions for the role of ERP systems in business. But precisely because of this, the need for strategy increases – not decreases. Organisations that only chase the latest technology risk investing in platforms without clear business value. Organisations that wait too long risk locking themselves into a technical environment that becomes increasingly difficult to develop.
The crucial thing is that each organisation formulates its own way forward. The strategy needs to balance business value, ability to change, technical debt, data availability, vendor development and risk. Only then can the question of the ERP system’s future be answered in a way that is not governed by the pace of technology, but by the business’s long-term needs.
Read the report "The State of ERP and Agentic AI", where we analyse how AI agents, automation and new technological possibilities affect the ERP market and what consequences this has for organisations' future system choices.
