The Next Era of ERP Will Look Nothing Like the Last One

This article, which expands on insights from a recent episode of the Evolving ERP podcast, delves into the current, transitory state of the ERP marketplace and why this next era will look different than the ones that came before.

Enterprise resource planning has spent three decades as the operational backbone of global commerce. What comes next will require a fundamentally different kind of platform than the ERP marketplace was built on. As can be expected, then, ERP is in the middle of a transition. The platforms that defined enterprise software from the 1990s through the 2010s are under pressure from every direction: hyperscalers encroaching from the infrastructure layer, purpose-built vertical software taking share from the top, and AI shifting what enterprise buyers actually need from a system of record. Understanding where ERP goes from here requires understanding how thoroughly the assumptions of the last era have broken down.

Quick Reference

What this article covers: Why the assumptions that built the ERP category no longer hold, where the gaps between current platforms and modern enterprise needs are widest, and what the next generation of ERP will have to look like to remain relevant.

Core claim: More than 70 percent of recently implemented ERP initiatives will fail to fully meet their original business use-case goals, according to Gartner, and many industry professionals also claim these systems address only a fraction of the modern enterprise’s operational footprint.

Who should read this: ERP buyers and evaluators, enterprise architects, CIOs navigating legacy platform decisions, and technology analysts tracking the competitive dynamics between incumbent suite vendors and emerging point solutions.

Key terms:

• Clean core: SAP’s architectural philosophy of reducing custom code to enable cloud migration; in practice, the “core” represents a shrinking share of enterprise operations.
• Composable architecture: Assembling enterprise functionality from modular, interchangeable services rather than monolithic suites.
• Ring-fencing: The practice of surrounding a legacy ERP core with best-of-breed cloud applications, which has proliferated far beyond what most organizations anticipated.
• Edge applications: Purpose-built solutions addressing operational requirements that emerged in the last five years and that ERP platforms were not designed to handle.

The Shrinking Core

When Gartner coined the term “ERP” in the early 1990s, the expectation was that these platforms would eventually cover the majority of enterprise operations. The logic was sound at the time: a unified data model, shared processes, and a single source of truth would naturally expand to absorb more and more of what businesses actually did.

That expansion never happened at the scale anyone predicted. Today, depending on the industry vertical, ERP systems can struggle to handle the majority of systems that a modern enterprise actually needs to run. The rest has been absorbed by specialized platforms, built internally, or left unaddressed.

This reflects a genuine mismatch between how ERP platforms were designed and how the economy actually operates. ERP emerged from manufacturing, specifically from MRP and MRP II lineages built for discrete production environments with bills of material, work orders, and inventory tracking as the organizing logic. That heritage is still visible in how these platforms think about processes, data structures, and integration patterns.

The problem is that the manufacturing model that ERP systems were designed to represent is a fraction of the global economy. Utilities need usage capture, billing, and capital-intensive asset management that ERP does not provide out of the box. Banks carry books of record that have no analog in a manufacturing data model. Insurance companies run on claims processing and policy management. Healthcare organizations now depend on telemedicine infrastructure that became mission-critical during the COVID-19 pandemic and has not retreated since. Each of these sectors has unique statutory reporting requirements, tax structures, e-invoicing mandates, and customs compliance obligations that vary by country and that multiply as emerging markets grow faster than either Europe or the United States.

The vendors have acknowledged these gaps in varying degrees, but acknowledgment has not translated into meaningful investment. Saying you serve the service sector because your customers use your financial and HR capabilities is not the same as building software that reflects how those industries actually operate. The distinction matters, and buyers are increasingly aware of it.

What Got Built Instead

The gap between what ERP promised and what it delivered was not left empty. Several technology categories emerged specifically because ERP did not evolve quickly enough.

Digital marketing and CRM provide the clearest example. While ERP vendors had the relationships, the transactional data, and the installed base to own customer engagement software, Salesforce built an entirely separate ecosystem in the CRM space, and Google and Meta captured digital advertising in ways that ERP platforms never seriously pursued. The data that would have made ERP vendors formidable competitors in marketing technology sat inside their own systems and went largely unused for that purpose.

The industrial Internet of Things followed a similar pattern. Manufacturing customers were already running ERP systems, but when industrial IoT began generating operational data at scale, it was Siemens, GE, and other large industrial technology companies that moved to monetize that data stream. ERP vendors might’ve had the customer relationships, but they lacked the capital appetite or architectural ambition to build the required infrastructure. The result is that some of the most valuable operational data in manufacturing now flows through platforms that lack deep integration with the ERP systems those manufacturers also run.

Cloud infrastructure is the sharpest example of the investment gap. Microsoft Azure, Amazon Web Services, and Google Cloud represent the defining technology infrastructure of this era. Building cloud infrastructure requires sustained, significant capital expenditure on data centers, networking, and GPU density. ERP vendors have historically been allergic to that kind of capital commitment. The consequence is that they now run their own SaaS products on infrastructure they do not control, while the hyperscalers who do control that infrastructure are increasingly entering enterprise application markets directly.

The Automation Gap on the Shop Floor

One of the more revealing mismatches between ERP ambition and industrial reality involves automation. Modern distribution and manufacturing environments have transformed operationally in ways that legacy ERP architectures were not built to support.

Consider what a contemporary large-scale distribution center actually runs on: autonomous mobile robots handling pick-and-place operations, conveyor systems with dynamic routing logic, machine vision for quality inspection, sensors feeding real-time OEE calculations back into production planning systems, and 5G-enabled wireless communication across the plant floor that eliminates the infrastructure constraints of wired networks. The warehouse management systems embedded in ERP suites were designed for human-operated environments with barcode scanners and periodic cycle counts. They are not wrong, exactly, but they are not sufficient for environments where the operational model has been rebuilt around robotics and automated material handling.

The same observation applies to manufacturing execution. MES integration, real-time machine time reporting, variance analysis from sensor data, and OEE calculation are technically achievable with current technology, but achieving them requires connecting ERP systems to plant-floor systems in ways most ERP implementations never prioritize. The data is available. The connections are possible. Implementations rarely reach the depth needed to make ERP genuinely useful as an operational intelligence platform rather than a financial record-keeping system with manufacturing modules bolted on.

This points to something structural in how ERP implementations have traditionally been approached. The dominant model has always been white-collar automation: finance, HR, procurement, and reporting. The operational core of manufacturing, logistics, and distribution has been treated as adjacent rather than central. That prioritization made sense in 1995. In a manufacturing environment where significant portions of direct labor have been replaced by automation, it represents a significant misalignment between where value is created and where ERP is actually deployed.

The Cloud Migration Stall

The major ERP vendors have spent the last decade pushing their installed bases toward cloud deployments. The results have been considerably slower than the vendors publicly anticipated. Rough estimates suggest that a not-insignificant percentage of SAP’s installed base has not yet begun a cloud migration.

The reasons customers give for staying on-premises are worth taking seriously. Migration costs are real and significant. Legacy ERP customizations, accumulated over decades of implementation work, represent a complicated technical debt that clean core architectures do not eliminate so much as externalize. Organizations that have rationalized those customizations as stable, depreciated investments are not wrong to view the cost-benefit calculation skeptically.

There is also a human capital dimension that migration advocates can underestimate. Many organizations have experienced staff who deeply understand their legacy ERP environments, and many of those individuals are not yet at retirement age. The institutional knowledge embedded in existing implementations has genuine value, and the argument that younger workers will not want to maintain these systems is less compelling when the people who actually know the systems are still in the building.

The third-party support market has emerged as a significant countervailing force. Vendors have built businesses around providing continued support for legacy ERP versions at meaningful cost reductions compared to vendor support fees, and they have been willing to commit to multi-year support guarantees. For organizations that want to extend the life of their on-premise investments while directing discretionary technology budget toward higher-priority modernization areas, this is a rational option.

The practical implication for organizations navigating this decision is that the on-premise-to-cloud timeline is more flexible than vendor roadmaps suggest. Using that flexibility strategically, specifically to invest in smart factory capabilities, IoT integration, and shop-floor automation in the interim, may generate more near-term value than a cloud migration that primarily reproduces existing functionality in a new deployment model.

The Data Ownership Question

A less-discussed but consequential dynamic is emerging around data ownership. ERP systems contain some of the most strategically valuable data in any organization, including transactional history, supplier performance data, production records, and customer behavior patterns. As AI applications begin to derive substantial value from large, clean operational datasets, the question of who controls that data and who benefits from training AI models on it is becoming commercially significant.

Some enterprise customers are taking explicit positions. Organizations in the pharmaceutical industry, for example, hold molecular data with enormous commercial value. Oil and gas companies hold seismic data that competitors would pay significantly to access. These organizations are not simply resistant to cloud migration on cost grounds; they are making deliberate decisions about data governance that reflect an understanding that their operational data is itself a strategic asset.

The legal dimensions of this dynamic are still developing. The litigation between major media organizations and AI training platforms over content licensing has established that this is not merely a theoretical concern. Enterprise customers who understand the value of their operational data are beginning to think about it the same way, and ERP vendors who assume that SaaS migration automatically includes rights to use customer data for model training are likely to encounter significant pushback.

The Composable Question

Composable architecture has been discussed as the answer to ERP’s structural limitations for several years. The basic proposition is that assembling enterprise functionality from discrete, interoperable services offers greater flexibility than monolithic suites, enables organizations to adopt best-of-breed capabilities without wholesale platform replacement, and creates a technology landscape that can adapt more quickly to new operational requirements.

The concept is sound, but the ecosystem is not yet mature enough to be a complete solution. A composable architecture is only as useful as the pool of available services, and that pool currently skews heavily toward white-collar functions: project management, HR workflows, financial approvals, and procurement automation. The operational capabilities that ERP currently provides, however imperfectly, do not yet have robust composable equivalents across the range of industries and processes where ERP is deployed.

This will change. The trajectory toward more modular, API-first enterprise software is durable. But organizations considering composable approaches today should be realistic about the complexity of integrating many point solutions, the data integration challenges that arise when operational data is distributed across dozens of systems, and the governance burden of maintaining a heterogeneous technology landscape.

What the Next Era Actually Requires

The shape of the next ERP era is coming into focus, even if most incumbent vendors have not yet committed to building it. Four areas will define whether the category reinvents itself or continues to shrink relative to the broader enterprise technology landscape.

Vertical depth over horizontal breadth.

The attempt to serve every industry with a common core supplemented by industry-specific modules has produced platforms that are too shallow to meet the specific requirements of utilities, financial services, healthcare, and other sectors that represent the majority of global economic activity. Real investment in vertical functionality, including regulatory compliance, industry-specific data models, and workflows designed for how those industries actually operate, is necessary.

Operational integration, not just operational reporting.

Connecting ERP to plant-floor automation, IoT sensor networks, and robotics systems as a first-class integration concern rather than an afterthought would reposition these platforms as genuinely useful for the operational core of manufacturing businesses. This requires architectural changes, not just new connectors.

Global localization at scale.

The emerging market opportunity is real and growing faster than the Western markets where ERP vendors have historically concentrated. Supporting the statutory reporting requirements, tax structures, e-invoicing mandates, and language requirements of markets in Southeast Asia, Latin America, and Africa is not a nice-to-have for vendors with global ambitions.

Implementation automation.

ERP implementations have been labor-intensive for 30 years, despite the accumulated knowledge from hundreds of thousands of completed projects. Data conversion tools, testing automation, configuration automation, and AI-assisted implementation support could substantially reduce the cost and risk of ERP projects. The fact that this has not happened at scale reflects incentive structures in the implementation partner ecosystem rather than any technical barrier.

None of these shifts is an incremental improvement to the existing model. They require ERP vendors to make investments, accept architectural disruption, and enter markets where they have no established advantage. That is precisely what the leaders of the last era avoided doing, and it is why the next era will belong to whoever is willing to do it differently.

This article draws on themes discussed in the Evolving ERP podcast on the Insight Jam network, featuring ERP practitioner Dan Aldridge and industry analyst Vinnie Mirchandani. You can find the full episode here.