The problem everyone knows about but nobody solves
Every BC implementation reaches a point where the consultants declare the configuration "done" and everyone prepares for go-live. And every experienced BC professional knows what happens next: the first week in production is spent firefighting errors that should have been caught before users ever touched the system.
The posting group combination that nobody created because it didn't appear in the test scripts. The number series that's three invoices away from exhaustion. The dimension set to "Code Mandatory" on the customer card but nobody assigned a default value. The vendor posting group that points to a GL account somebody blocked last month.
These aren't complex problems. They're configuration completeness problems — and they're entirely preventable with a systematic check.
What a proper go-live validation looks like
A comprehensive BC configuration validation needs to check six areas:
1. Posting group matrix completeness
Every combination of General Business Posting Group × General Product Posting Group that appears on any master data record must have a corresponding row in General Posting Setup with populated Sales and Purchase accounts. Miss one combination and the first transaction that hits it will fail with a cryptic error message.
2. Number series health
Series that are 90%+ exhausted, series with expired date ranges, setup pages with blank number series fields. These are ticking time bombs — they work fine until suddenly they don't.
3. Dimension configuration
Mandatory dimensions without default values on master data. Blocked dimension values still assigned as defaults. These cause posting failures that are particularly hard to diagnose because the error message rarely says "your dimension is blocked."
4. Setup table completeness
The GL Setup, Sales Setup, Purchases Setup, and Company Information pages have hundreds of fields between them. The setup wizard doesn't cover all of them. How many blank fields are lurking in yours?
5. Chart of accounts structure
GL accounts referenced in posting setup that are blocked or have Direct Posting disabled. Posting accounts without Account Categories, which causes gaps in financial reports.
6. Master data readiness
Customers without posting groups. Vendors without payment terms. Items without a Base Unit of Measure. These records were probably created during data migration and never fully completed — they pass creation validation but fail at transaction time.
The manual approach (and why it doesn't scale)
Most partners maintain personal checklists in Excel or OneNote. Some have built their own AL reports that check specific tables. The problems with this approach:
- Inconsistent. Every consultant's checklist is different.
- Incomplete. Checklists drift out of date as BC evolves.
- Not repeatable. You can't run an Excel checklist against live data automatically.
- Labour-intensive. Checking 240 posting group combinations by hand isn't practical.
An automated alternative
XPT Validate runs a systematic configuration audit against a live BC environment in seconds. It connects through the same API infrastructure as XPT Tools — no additional extensions to install — and checks all six areas described above against a comprehensive rule set.
Every finding is categorised as Critical (will break), Warning (will cause pain), or Info (best practice gap). The results are a clear, prioritised list that tells a consultant exactly what to fix before go-live.
After go-live, Validate becomes an ongoing data quality tool: user-defined validation templates check master data as it's created and maintained, with a quality score dashboard that tracks your environment's health over time.
What you'll typically find
We've run XPT Validate against dozens of BC environments. Here's what we typically find:
- 5–15 posting group gaps on a mid-sized implementation
- 8–12 exhausted or near-exhausted number series (especially on systems that have been running for a while)
- 2–5 mandatory dimensions without defaults (the most common cause of "unexpected" posting errors)
- 10–30 master data records with missing posting groups (usually from bulk data imports)
On one corporate sandbox, we found 230 critical issues in 5 seconds. Every one of them would have surfaced as a user-facing error within the first month.