/Posts
D365 configuration drift: what it looks like after go-live
16 November 2026

ERP Guardians

- D365 configuration drift
- Microsoft Dynamics 365
D365 configuration drift is the gradual movement of settings such as payment terms, tax groups, number sequences and posting profiles away from the agreed design, or out of alignment across environments and legal entities. It can develop through urgent fixes, refreshes and local exceptions. Detecting it starts with a scheduled comparison against a named baseline, followed by a review of each difference.
What is D365 configuration drift?
Dynamics 365 Finance and Supply Chain Management (D365 F&O) is configured through thousands of settings. At go-live, the approved design provides the reference point for those settings. As the live system evolves, that reference needs to remain aligned with the changes the business approves.
Configuration drift is the gap that can open between the approved design and the live system. It often develops gradually. A posting profile is corrected during month-end. A parameter is adjusted to meet an urgent need in one legal entity. Each change may be reasonable, but without an updated record, it becomes harder to establish which configuration is intended.
Drift appears in three forms:
- Environment drift: UAT, production and other environments hold different values for the same setting.
- Legal entity drift: companies that should follow one standard end up with different payment terms or tax settings.
- Design drift: the live value no longer matches what the approved design says, even if every environment agrees with each other.
Design drift can be harder to spot: two environments that have changed in the same way will still match each other.
How does configuration drift happen after go-live?
Configuration drift often develops through practical responses to day-to-day needs. Five common sources are worth reviewing.
- Changes typed straight into production. An urgent issue is corrected in the live system, but the change may not be documented or applied in the other environments where it is needed.
- Fixes made in one environment only. A tester corrects a setting in UAT so a scenario can proceed. The correction addresses the issue, but is not carried through to production.
- Database refreshes. A refresh copies the database from one environment to another. Microsoft documents that the refresh from production to a sandbox does not carry over every item, for example email addresses and print management settings, so a refreshed sandbox is similar to production but not identical to it.
- Per-entity exceptions. A subsidiary needs a different payment term or number sequence. The change is agreed and configured, but the reason for the exception is not recorded alongside the approved design.
- Service updates and new features. Sandbox and production environments may be at different stages of an update or feature rollout. Where a feature is enabled in one environment but not another, behaviour can differ even without a direct change to the underlying business settings.
The governance gap is often the same: an incomplete record of what changed, why it changed and who owns the decision.
Why is a difference not automatically a defect?
A comparison between two environments may reveal differences. Some are intentional and appropriate.
A test environment should use an appropriate test email configuration. A legal entity in another country may legitimately need a different tax group. Treating every difference as an error adds noise to the review and makes it harder to focus on findings that need action.
An expected difference has a documented reason and an owner. The values still differ, but the business understands why. The review therefore needs to establish not just whether the settings match, but whether any difference has been agreed.
- Email server or print settings in a test environment: Typically expected?: Yes. Why: Test systems should not send real mail.
- Different sales tax group between two countries: Typically expected?: Often. Why: Local law and local design.
- Credit limit type differs between UAT and production: Typically expected?: No. Why: The tested control is not the live control.
- A feature switched on in one environment only: Typically expected?: Rarely. Why: Behaviour will differ in testing and live use.
Table 1. Whether a difference is expected depends on whether anyone decided it.
What does drift look like? A worked example
Consider a fictional distributor with three legal entities and two environments: UAT and production. The design approved at go-live specifies the same payment terms and credit limit type across all three entities, with different tax groups only where country requirements justify them.
Eight months after go-live, a comparison against the approved design would show the following.
- Payment terms: one entity has a new term, "Net 45", added by a finance user for a negotiation.
- Tax groups: a tax group was edited in production to fix a rounding complaint. UAT holds the old definition.
- Number sequences: a sequence in one entity was set to continuous after an audit remark. The other two entities are not continuous.
- Posting profiles: a customer posting profile was changed during a month-end and left in place.
- Credit limit type: UAT was corrected during testing. Production was never updated.
Two findings may be deliberate (the new term and the continuous sequence) and need their rationale recorded. Two may be unintended (the posting profile and the credit limit type) and need review before an approved correction. The tax group edit needs discussion, because the approved design may no longer reflect the business requirement.
The review is about understanding and classifying differences, not assigning blame. Making those differences visible gives the team a clear starting point.

Typical sources of drift: each one moves a setting without updating the record of what was agreed.
How do you detect D365 configuration drift?
Detection needs three elements: an agreed reference point, a defined scope and a regular review schedule.
- Name a baseline. A baseline is an approved state of the settings that matter, with a name, a date and an approver. It gives reviewers a shared reference for deciding which differences need attention.
- Choose the scope. Start with settings that carry financial and compliance risk. Include the legal entity, because the same setting can hold different values in each company.
- Compare at setting level. Show the value on each side, so reviewers see that a setting is "Balance" here and "Balance + all" there.
- Record the expected differences first. This reduces repeat investigation and helps reviewers focus on unexplained changes.
- Repeat on a schedule. A comparison before a release shows the configuration at one point in time. Regular checks make subsequent changes visible, with a named owner responsible for reviewing the results.
A spreadsheet of exported values captures a point in time. It remains useful as a baseline when its approval, scope and subsequent revisions are managed.
What should you do with a finding?
Each finding should lead to a documented decision. There are three main outcomes.
- Accept it as an expected difference. The difference is appropriate. Record the reason and owner, and set a review date so that the exception remains relevant as requirements change.
- Correct it through an approved change. The difference is unintended. Restore the baseline value through a documented change requested by one person and approved by another, keeping the correction traceable.
- Update the baseline. The live value reflects an approved requirement that the design no longer captures. Approve the revised baseline so that the reference and the intended configuration align again.

Three outcomes for a finding: accept it, correct it, or update the baseline.
Whichever outcome applies, record the decision and ownership so the finding is not left unresolved.
How does Warden handle configuration drift?
Warden for Dynamics 365 is a configuration-governance product for D365 F&O, built to support this review process.
In Warden, a snapshot of the agreed settings provides the baseline. A monitor compares a defined set of settings in one legal entity against that baseline on your chosen schedule. The Monitors page shows whether monitoring is enabled, who receives alerts and whether the environment is in sync.

Screenshot from a test environment: a monitor for a group of customer settings, with monitoring switched on (1), the overseer team that receives alerts (2), the drift status, here Drift Detected (3), and the Scan for Drift command (4). The baseline snapshot and the open alert are on the right.
When live settings differ from the baseline, Warden raises a drift alert. It groups modified, missing and added rows by area, such as terms of payment, and displays the baseline and environment values side by side. The reviewer can restore selected rows to the baseline, accept the environment values as the new baseline, or ignore the alert. Restore and accept require a second person only when approval is enabled for the environment. In that case, the requester and approver must be different people.

Screenshot from a test environment: a drift alert with one tab per area that drifted: the count of modified, missing and added rows (1), the switch between field labels and technical names (2), and the restore (3), accept (4) and ignore (5) actions.
Approved differences can be recorded as expected differences to exclude them from later alerts. Each action leaves a record of who acted and when, supporting subsequent reviews.
Frequently asked questions
What is configuration drift in Dynamics 365?
Configuration drift in Dynamics 365 is a gap between the agreed configuration and the live settings, or between environments and legal entities intended to align. It can develop through individually reasonable changes that are not documented or applied consistently where needed.
How do I compare D365 environments?
Define the settings in scope, choose the reference environment and compare values at setting level for each legal entity. Record expected differences, such as test email configurations, so reviewers can focus on unexplained changes. Repeat the comparison on a schedule to identify changes over time.
Is every configuration difference between environments a problem?
No. Differences such as email configurations, bank details or country-specific tax groups may be intentional. An unexplained difference needs review, rather than an automatic correction. Document each intended difference with a reason and an owner to make that distinction clear.
Does a database refresh make environments identical?
Not completely. Microsoft documents that a refresh from production to a sandbox does not copy certain items, including email addresses, print management settings and the status of users other than the admin. Environments can therefore differ right after a refresh, and they continue to diverge as people change settings afterwards.
How often should we check for configuration drift?
Use a fixed schedule that supports review before key releases or period-end activities. The appropriate interval depends on how frequently settings change and how many people can change them. Assign a named owner to receive and review each result.
Where to start
Start with one legal entity and a focused scope, such as payment terms and tax groups. Agree a baseline, document the intended differences and run the first comparison. Allow time in the initial review to establish which findings need action.
For Microsoft background, the Learn pages on refreshing a database and on service updates describe how environments are copied and updated.
To explore this approach using your own scenario, message us for a look at D365 configuration drift monitoring with Warden.