The statistics are well-known and consistently grim. More than 70% of large-scale enterprise integration programmes — ERP implementations, cloud migrations, CRM rollouts — run over budget, miss deadlines, or fail to deliver their intended business outcomes. In the Asia-Pacific region, where organisations are navigating the dual pressures of rapid digital adoption and legacy infrastructure debt, the failure rate is no better.
What is striking, given the billions spent on these programmes annually, is how consistent the failure modes are. The root causes are rarely technical. The technology mostly works. What fails is everything surrounding the technology: governance, change management, data quality, and the organisational dynamics that determine whether a transformation programme is set up to succeed or doomed before the first line of code is written.
The Five Failure Patterns
-
01
Scope Defined by Vendor, Not BusinessWhen scope is defined primarily by the implementing vendor rather than the business, it tends to reflect what the vendor knows how to build rather than what the organisation actually needs. Requirements workshops become feature demonstrations. The business signs off on a scope it does not fully understand and discovers the gaps during UAT.
-
02
Data Quality Treated as Someone Else's ProblemIntegration projects expose data quality problems that organisations have been papering over for years. Duplicate customer records, inconsistent product codes, unmaintained master data — these surface during migration and consume enormous amounts of time that was not budgeted.
-
03
Insufficient Business OwnershipWhen integration projects are owned primarily by IT, business stakeholders disengage. The handover moment — when IT declares the system ready and hands it to the business — is often when the real problems surface.
-
04
Change Management as an AfterthoughtChange management is frequently scoped as a small workstream to satisfy a procurement requirement rather than a programme-critical activity. Organisations that invest seriously in change management achieve significantly higher adoption rates and return on investment.
-
05
Architecture Designed for Go-Live, Not the FuturePoint-to-point integrations that solve the immediate problem but create a fragile web of dependencies are one of the most common sources of long-term technical debt. Organisations that invest in an integration platform upfront pay more at the start but spend dramatically less on maintenance over the following five years.
What Successful Programmes Do Differently
The organisations that consistently deliver successful integration programmes share a number of characteristics. They assign senior business leaders with genuine decision-making authority as programme sponsors. They conduct independent assessments of their data quality before committing to a go-live date. And they build internal capability alongside vendor delivery, so the organisation is not wholly dependent on external resources after go-live.
"Every integration project we've seen succeed had one thing in common: a business owner who treated it like their programme, not IT's programme." — JH3 Principal Consultant
They also resist the pressure to go live on a fixed date when readiness indicators suggest the organisation is not ready. The cost of a delayed go-live is highly visible. The cost of a failed go-live — loss of data integrity, user rejection, operational disruption — is often catastrophic and far exceeds the cost of a measured delay.
The good news is that these failure patterns are well understood and preventable. With the right governance structure, adequate data preparation, and genuine business ownership, enterprise integration programmes can and do deliver their intended value. The discipline required is organisational, not technical.