Organizations spend months, sometimes years, building up to an ERP go-live. They define the technology, select the software, redesign processes and configure the system. Then comes a step that rarely gets the attention it deserves: user acceptance testing. UAT is where the rubber meets the road. It’s the first time your design, your data and your integrations all meet the people who will actually run the business on the new system.
Too often, though, user acceptance testing for ERP is treated as a formality, a set of scripts to run before the real work of go-live begins. That’s a mistake that can shut down operations. This article covers what separates UAT that protects the business from UAT that just checks boxes, drawing on a recent conversation with my colleague Julie Granshaw, a former supply chain director who now helps our clients with business readiness.
Table of Contents
ToggleWhat Is User Acceptance Testing in an ERP Project?
User acceptance testing is the phase where the business, not the vendor or system integrator, confirms that the new system supports how the organization actually works. Earlier testing phases, like unit, system and integration testing, prove that the software is configured correctly and that data flows between systems. You can learn more about those phases in our guide to system, integration and user acceptance testing.
UAT answers a different question. It’s not “does the software work?” It’s “can we run our business on it?” Those sound similar, but in our experience, the gap between them is where many ERP go-lives fall apart.
Checkbox UAT vs. UAT That Tests the Business
Most system integrators hand clients a set of standardized test scripts. In SAP’s world, that means scenarios like order to cash, procure to pay and record to report. A tester runs through the script, confirms each step works and signs off. That’s valuable because it proves the configuration and the data for that scenario are correct.
What standard scripts don’t do is account for what makes your business different. They aren’t designed around your unique processes or the daily challenges your teams actually run into. The one-off transactions, exceptions and edge cases that standard scenarios skip can be the very things that stop operations on day one.
Here’s the disconnect we see over and over. The vendor or integrator says the software works, they’ve tested it against the processes and the data is flowing. Technically, they’re often right. Then the switch flips and operations grind to a halt, because the software works but it doesn’t work for the business.
Who Should Perform User Acceptance Testing?
The people doing the job should be doing the testing. In practice, that means a team of super users or key users, one or more for each functional area in scope. The ideal tester sits in a specific spot in the organization:
- Senior enough to understand every process and responsibility in their functional area, not just one or two transactions.
- Hands-on enough to know how the work really gets done in the system day to day.
- Invested in the project’s success, with real skin in the game.
- Given the time to test, rather than squeezing UAT around a full-time day job.
That last point is where many projects fall down. When senior people are busy elsewhere, organizations send a junior employee or a new hire to “run the scripts.” They’ll follow the steps, but they won’t know what to look for, and that gives leadership false confidence going into go-live.
How to Write UAT Test Scenarios That Reflect How You Really Work
Good UAT test scripts reflect how the business actually runs, not just how the software is designed to work. Here’s how to get there.
Start With Your Requirements
Go back to the requirements you gave the vendor during design and configuration. For each one, ask what the software addressed, where the gray areas are and whether your UAT plan accounts for it. Requirements that don’t fit a standard scenario still have to work somehow in the system, and they need their own documented test.
Capture a Real Week of Work
Ask each key user to document every type of transaction they handle in a typical week. It’s a big ask that takes hours, but it surfaces the critical tasks that standard scripts miss. Processes will change with the new system, so how you get to an output may be different in the future. What those outputs are, and the level of detail they require, should still be captured.
This work builds on what you did in design. If your current-state processes were never documented, UAT is where that gap shows up. That’s why as-is process mapping pays off long before testing begins.
Don’t Fear a Long List
One client had 35 test plans. Another had 58, and the team felt overwhelmed. That’s normal, and it’s better to have too much detail than not enough. AI tools can also help here by turning requirements and design-session notes into draft test scenarios, especially for requirements that fall outside standard business flows. Key users still need to review and complete them.
How to Prioritize UAT When Time and People Are Limited
Most projects face resource constraints during UAT. Business process owners and key users still have businesses to run, and few are allocated to testing full-time. The answer isn’t to test less. It’s to test smarter by ranking every scenario before testing begins.
| UAT Scenario Priority | Testing Goal | If Testing Can’t Be Completed |
|---|---|---|
| Critical | Test and pass every scenario. | Raise it as a major project risk and don’t go live. |
| Medium | Complete the majority. | Escalate to leadership with a documented risk. |
| Low | Test as time allows. | Document a workaround and plan to resolve it during hypercare. |
If anything above low priority can’t be tested, flag it in your weekly status reports and steering committee meetings. Leadership needs to know there’s a real risk of issues at go-live. The best fix is to assign key users to UAT full-time and backfill their day jobs, not to hand the scripts to whoever is available.
Record every decision about what you’re willing to risk in a key decision tracker. That includes what you’re conceding, why and how you’ll handle the fallout. Those records become critical inputs to your go/no-go decision.
Why UAT Problems Usually Start Months Earlier
When we walk into a project, one of the first things we look at is how involved leadership is. UAT exposes the gaps left by earlier phases. These are the most common:
- The wrong people in design workshops. When senior staff skip sessions and send junior stand-ins, unique requirements never make it onto the list. They surface in UAT as “nobody ever mentioned that before.”
- Unrealistic plans. A plan that assumes one 30-day UAT cycle with a single revision often turns into 90 days once testers see how much there is to cover.
- Unfinished integrations. Testers are told to work around integrations that are still being fixed, so true end-to-end testing never happens.
- Unclear ownership. Statements of work typically assign UAT to the customer. The integrator reports status only on what it owns, so UAT gaps can hide behind a green status report.
That last point is one of many responsibilities that fall outside a vendor’s scope. We cover the rest in our article on what your ERP system integrator can’t do.
The Bottom Line
User acceptance testing is the moment when every decision you’ve made about design, data, integrations and people comes together. Whether it works shows how well you prepared. Treat UAT as a formality and you’ll find your gaps after go-live, when they’re far more expensive to fix. Treat it as a true test of your business, with the right people, realistic scenarios and clear priorities, and it becomes your best protection against a disruptive launch.
For more on what to test and when, see our guide on how to test software before go-live.
Eric is recognized globally as a leading voice in digital transformation and ERP strategy. Over the past three decades, he has helped hundreds of organizations, including Nucor Steel, Fisher & Paykel Healthcare, Kodak, Coors, Boeing, and Duke Energy, define their technology roadmaps, modernize complex operations, and deliver real business value from large-scale transformation initiatives.
As Founder and CEO of Third Stage Consulting, Eric leads an independent, technology-agnostic advisory firm focused on helping clients navigate the shift from traditional ERP to more flexible, AI-enabled Digital Enterprise Operations (DEO) models. His work spans ERP selection, implementation quality assurance, organizational change, and operating model design across a wide range of industries and geographies.
Eric is also a prolific thought leader, known for his pragmatic takes on AI, cloud, and enterprise software trends, as well as his firm’s benchmark research and frameworks for de-risking transformation. He is dedicated to helping executive teams cut through vendor hype, make confident investment decisions, and successfully reach the “third stage” of their digital evolution.