Why ERP initiatives struggle to deliver business value

Changing ERP and rising expectations

Two major shifts are at play in the changing ERP landscape, according to Jonathan LaCour, CTO at Mission Cloud, a CDW Company. "First, the move to cloud-based ERP has shortened implementation timelines while raising expectations for continuous value delivery," LaCour said. "Second, ERP scope has expanded dramatically."

Ultimately, both shifts move the focus from implementation success to outcome realization. This shift reflects a larger challenge: managing change across enterprise systems to deliver predictable business value. While ERP is central to this issue, it is only one part of an evolving and increasingly complex business system.

"ERP programs used to chase a defined end state; today, they have to deliver value while the end state is still moving," said John Cook, director at global technology research and advisory firm Information Services Group.

Cloud ERP, hybrid architecture, frequent vendor updates, ecosystem integrations, regulatory change, supply-chain volatility and shifting workforce expectations all introduce more moving parts, Cook said. At the same time, companies "expect ERP to enable transformation while the business continues to change during the project." The result, says Cook, is that "value assumptions made at the start of the program can become outdated before the implementation is completed."

Where business cases break down

ERP strategies historically targeted limited objectives. "Our research shows companies frequently focus too narrowly on replacing legacy systems or improving isolated processes instead of reimagining end-to-end value streams and enterprise operations," said Colapietro.

On the other hand, ERP business cases were often too loosely defined. Business cases were "always counterfactual fictions," said Aman Mahapatra, chief AI strategy officer at tech consulting firm, Tribeca Softech. "The promised ROI was calculated against a 'what would have happened without the ERP' baseline that nobody measured at the time and nobody can reconstruct afterward," Mahapatra added.

Although current numbers might appear more objective, they can still be misleading.  

"The savings that show up in year-three reporting are largely accounting artifacts: depreciation schedules favorable to the new system, head count reductions that would have happened anyway, and cost reallocations that move expense lines without changing total spend," said Mahapatra. He said that the CFOs he works with at large financial institutions "have stopped pretending these numbers are meaningful."

The lack of follow-through continuity widens the gap between a functioning ERP system and realized outcomes. "Nobody who was in the room when the contract was signed is accountable for the realization phase," said Mahapatra.

In the typical scenario following implementation, Mahapatra says the internal program manager has moved on; the executive sponsor has changed roles; the system integrator partner has rotated its A-team to a new client; and the business unit owners are managing operations rather than chasing the original business case.

"The gap between technical success and business failure is the gap that this incentive structure creates, and no amount of better project management closes it," Mahapatra added.

The gap between technical success and business failure is the gap that this incentive structure creates, and no amount of better project management closes it.

Key complications can stem from more nuanced causes, too. An ERP implementation changes how people work, who has authority over what data, and which department owns which decision, according to Frank Meltke, CEO of contraco Management Consulting, a global digital transformation consultancy. "Those are political questions dressed in technical clothing," he said. "The system integrator solves the technical problem. Nobody solves the political problem. Three years later, the system runs perfectly and the organization works around it."

Ultimately, delivering business outcomes means rethinking how business is done rather than how it is recorded. If an organization has not yet "redesigned processes, changed behaviors, improved data discipline or aligned incentives around the new way of working," said Cook, "ERP becomes a new platform for old habits."

Governance and accountability gaps

The accountability gap is structural. Consider that accountability is historically structured to cover every move up to and including the go-live date. "Three practices show up consistently," Mahapatra says, in the institutions he works with that realize ERP value, "and absence of any one of them predicts failure with depressing reliability." The first is treating go-live as the starting line rather than the finish line. The second is investing in primary data stewardship as a permanent capability rather than a one-time implementation activity. The third is refusing to customize the ERP itself and instead building the differentiating workflow logic in a layer above it.

But accountability also means assigning someone to be in charge and making them responsible for the ultimate outcome.

"Accountability ensures that business leaders, not just the program team or systems integrator, own the outcomes promised in the business case. Without that structure, value realization becomes everyone's aspiration and no one's responsibility," said Cook. He added that ERP benefits do not materialize because a steering committee approved them; they materialize "because accountable leaders make the operating changes required to achieve them."

Governance often separates ERP programs that merely install software from those that deliver business value. "Clear decision rights help prevent endless customization, unresolved process debates and local exceptions that dilute standardization," Cook said.

Complexity as the hidden variable

Complexity in ERP systems continues to grow. ERP initiatives today are far more complex because companies are no longer modernizing systems in isolation as they simultaneously try to prepare for AI-driven operating models, redesign workforce structures and create more connected enterprise ecosystems, according to Colapietro.

The pace of AI innovation is also quickening, making decisions about things like defining long-term architectures, governance models and predicting workforce impacts tricky to get right. The move toward open architectures and hybrid ecosystems, where ERP platforms must work in concert with multiple AI providers, orchestration layers and enterprise applications, Colapietro says, also adds to the ever-growing pile of complexity. And that's before you get to the long list of new expectations.

"Expectations have expanded dramatically. Companies now want ERP platforms to support predictive intelligence, intelligent automation, agent orchestration and enterprise-wide visibility, not just transactional processing," added Colapietro.

However, some complexities arise from past decisions, creating more obstacles than opportunities. The most common is the over-customization of ERP programs. Companies that stray too far from standardized, simplified processes often increase complexity, limit agility and make it more difficult to adopt future innovation, upgrades and embedded AI capabilities, according to Colapietro. "In many cases, customization preserves outdated ways of working instead of enabling modernization," she said.

But the deeper complexity problem "is not technical at all; it is human," said Meltke. The complexity that "kills ERP outcomes," he says, is the undocumented complexity: "the workarounds, the exceptions, the configurations nobody remembers why they made, the processes that exist because someone 20 years ago solved a problem nobody else knew existed." That complexity is insidious because it lives in human beings rather than in documentation, and while it does not disappear when those human beings retire, it does become invisible. "And invisible complexity in an ERP implementation surfaces at the worst possible moment, after go-live, in production, when the cost of failure is highest," Meltke said.

Separating value realization from underperformance

Ultimately, ERP provides structure that becomes the foundation of the business. Its performance can be measured but the metrics you select to use matter a great deal. Traditionally, ERP ROI is defined through hard metrics such as cost reduction, process efficiency, faster close cycles or lower technology maintenance costs.

"Increasingly, though, companies are broadening those metrics to include workforce productivity, resilience, agility, operational transparency and AI enablement. Where many ROI models still fall short is that they measure incremental efficiencies rather than transformational business impact," said Colapietro.

Measuring employee experience, decision-making and organizational agility also provides useful information. Ultimately, you're looking to measure more than the technical aspects of an ERP implementation.

Meltke said the following four things separate organizations that deliver from those that do not:

1. They define business ownership before technical ownership. A named business executive owns each major process domain and is accountable for outcomes, not just for system adoption.
2. They measure the right things before, during and after implementation. Not system uptime. Not go-live date. Business outcomes that existed in the original case for change.
3. They focus on transformation, not technology. They treat the implementation as an organizational change program that happens to include a technology component, not as a technology program that will eventually change the organization.
4. They do not declare victory at go-live. The real work begins the day the system is live. The organizations that succeed stay engaged with the business for 18 to 24 months after implementation, solving the problems that only become visible when real people use the system for real work.

In short, harvesting value from ERP is a job that is never done. Organizations that consistently realize ERP value, according to Cook, start with a small set of clear business outcomes and manage the program around those outcomes from design through post-go-live optimization. "They establish baselines, assign benefit owners, make tough process-standardization decisions, invest heavily in change management and continue improving after launch," he added.

Ultimately, the value of ERP is not determined at go-live; it is shaped by what happens afterward. The organizations that succeed are those that continue managing accountability, governance and process change long after the implementation team has moved on.

Pam Baker is a freelance journalist and the author of books including ChatGPT for Dummies and Generative AI for Dummies. Baker is also an instructor on AI topics for LinkedIn Learning and a member of the National Press Club, the Society of Professional Journalists and the Internet Press Guild.