Choosing where your ERP lives is one of the most consequential decisions you’ll make. Get it right and you’ll move faster, spend smarter, and keep teams aligned. Get it wrong and you’ll wrestle upgrades, workarounds, and mounting technical debt.
This guide breaks the choice down to its essentials. Costs, control, and speed, and closes with real-world lessons from a New Zealand organisation that moved from MYOB Greentree (on-prem) to a cloud platform.
Whether your ERP runs in your server room or in the cloud, you still need to do the hard, valuable work: map processes, clean data, engage stakeholders, and design for outcomes. The difference is how you pay for it, how you govern it, and how quickly you can evolve.
Cloud ERP delivers the application over the internet, typically as a subscription. The vendor operates the infrastructure, handles most updates, and provides APIs and ecosystems for extension.
On-prem ERP is installed on your own servers (or hosted privately). Your IT team manages infrastructure, patching, security hardening, and upgrade projects.
Cloud: predictable subscription (licences + platform resources) plus implementation and support. You avoid capex on hardware, data centre fees, and most infrastructure labour. Upgrades are included, which reduces the “big-bang” cost spikes that come every few years on-prem.
On-prem: perpetual (or term) licences + implementation plus servers, storage, backups, monitoring, disaster recovery, database licences, and the people to run it all. You’ll also budget for periodic major upgrades (testing, change windows, re-certifying integrations).
A useful mental model:
If your differentiation is how you operate (process, customer experience), you’ll likely get better ROI spending on implementation, change, and analytics—not infrastructure.
If your differentiation is a unique technical requirement (ultra-low latency on a plant network, extreme data residency constraints), the extra infrastructure cost may be justified.
Cloud gives you control where it counts, roles, workflows, approvals, integrations, data models, without owning the plumbing. Vendors provide role-based access, audit trails, encryption, regional hosting, and change windows you can plan around. You typically extend via configuration, low-code tools, and published APIs/SDKs.
On-prem gives you custody of the stack. You can tune the OS, database, and network, and time upgrades exactly to your calendar. That flexibility is powerful, yet it means you own patching, capacity planning, vulnerability management, and disaster recovery. Custom code can go deeper, but deep customisations also increase upgrade friction.
Key question: Where do you genuinely need sovereignty, on the infrastructure, or on the business rules and data? Most organisations want absolute control of process and data definitions, not of hypervisors and firmware.
Speed shows up in three places: time-to-value, time-to-upgrade, and time-to-integrate.
Cloud shines at all three. You stand up environments rapidly, consume frequent feature releases, and plug into ecosystems (connectors, iPaaS, marketplaces) to add capability without long build cycles.
On-prem can be fast with a mature internal platform team, but upgrades and integrations often turn into projects that compete with other IT priorities. “Windows” for change are narrower, and QA cycles are heavier.
For most mid-market and growing enterprises, cloud creates a shorter feedback loop: configure → test → adopt → measure → iterate.
Both models can be secure; the responsibilities differ.
Cloud: leading vendors operate to strict standards, with encryption, segregation, continuous monitoring, and regular audits. You retain responsibilities for identity, roles, data quality, and third-party add-ons.
On-prem: you can design security to your exact policies—but you must sustain it. Missed patches, weak network segmentation, or untested backups become your risk to carry.
If you operate under specific regulations, check data residency options, audit logging depth, retention policies, and disaster recovery objectives in writing—whichever path you choose.
Modern operating models are composable: a strong ERP core for finance and operations, surrounded by specialised apps (e-commerce, WMS, field service, planning & budgeting). Cloud ERPs typically offer clean REST APIs, webhooks, and established connectors; on-prem ERPs often support integration, but you’ll host and secure more middleware yourself.
The goal isn’t “one system to rule them all”, it’s a standardised core with well-managed edges.
You want lower IT overhead and predictable costs.
Multi-site or remote teams need secure access anywhere.
You value rapid iteration—frequent releases, faster pilots, quicker time-to-value.
You prefer configuration over custom code, and a robust marketplace of extensions.
You’re consolidating from several systems and want a clean, modern platform as part of the change.
You have hard constraints: air-gapped plants, unique latency needs, or strict policies that rule out public cloud.
You’re equipped to operate enterprise infrastructure (24×7 monitoring, patching, DR testing) and can fund periodic major upgrade projects.
You depend on deep, low-level customisations that a cloud platform won’t allow (and can quantify the value they deliver).
The Catholic Diocese of Auckland (CDA) operated multiple entities, over 50 schools and 70 parishes across a large geographic area, with server-based systems that didn’t integrate well. Finance teams spent significant time assembling board packs and consolidations from disparate data sources, with formats and reports differing across entities. After moving to MYOB Acumatica (cloud), they standardised reporting across entities, gained stronger audit trails, and reduced manual interventions, freeing time to focus on insight and long-term goals like interactive board dashboards.
Start with outcomes. Define measurable wins (e.g., faster month-end, standardised reporting, fewer stockouts).
Stair-step scope. Phase 1 for foundations (finance, core inventory, order processing), then expand to WMS, projects, manufacturing, planning & budgeting.
Design the data. Harmonise charts of accounts, item masters, customers, suppliers; agree on definitions before migration.
Keep the core clean. Configure before you customise; extend via APIs and certified apps.
Pilot with real processes. Use representative data (e.g., quote-to-cash, procure-to-pay) to validate fit and performance.
Invest in change. Communicate, train, appoint super-users, and run post-go-live “hypercare” with clear success metrics.
If your priority is speed of change and lower operational overhead, and you don’t have exceptional infrastructure constraints, cloud ERP is usually the pragmatic choice. If you must own the entire stack and can fund and sustain that ownership, on-prem remains viable.
At Verde, we help New Zealand organisations weigh the trade-offs in their real context, budget, risk, team capacity, and growth plans, and design a roadmap that delivers.
Keen to talk through your situation? Book a 30-minute ERP readiness chat and we’ll map the fastest, lowest-risk path.