IT systems · Practical guide

Excel co-authoring: testing simultaneous editing under GxP

Two people edit the same Excel workbook. Both see a plausible result while changes continue to save in the background. For a GxP decision, establish which version is complete, has been reviewed and is authoritative. Co-authoring changes how that evidence is produced.

Table of contents

At a glance

Can Excel co-authoring be used in a GxP process?

Co-authoring may suit a defined GxP workflow where clients, storage, identities, editing and approval rules are controlled and tested. Assess concurrent changes, interrupted synchronisation, historical traceability and protection of the authoritative record. A release calculation may require additional controls or a different division of work. AutoSave and version history alone do not demonstrate suitability.

First, define the common working process.

Microsoft describes co-authoring for supported Excel clients and workbooks on OneDrive or SharePoint Online. This is a technical requirement, not a GxP rating. In addition, distinguish the modern common editing of older share functions. Product name, file format, client version and used functions belong to the viewed status. A workbook checked on the desktop is not automatically usable in the browser.

Define the intended use of the shared file. A preliminary capacity planning has different requirements than a calculation supporting a quality-related decision. Describe inputs, calculations, reviewers, and subsequent usage. The critical transition typically occurs between "in common review" and "formally reviewed and approved." When no party is responsible for this transition, a recipient may treat an interim status as the final result. This process gap cannot be closed solely by storing the file in the cloud.

Sources: [1] [2]

Distinguish editing, saving and approval

AutoSave saves changes during work. This does not confirm the technical correctness of the changes nor constitutes a release. A user may inadvertently modify a formula, and the technically successful save operation preserves exactly that change. Therefore, define which fields are editable, which rules apply, and who confirms the finalization. Formula protection, access rights, and technical review serve distinct functions.

Describe a clear and unambiguous final status. This may involve a controlled version with restricted editing rights and documented decision-making. During evaluation, verify that subsequent changes are detected and addressed. A simple typed status "released" is ineffective if any user can subsequently modify values without the process responding. The technical implementation is verified against the actual Microsoft 365 environment and is not derived from a preferred workflow instruction.

Sources: [2] [1]

Parallelism alters the test design

A single-user test evaluates formulas and input rules, but does not assess the interaction among multiple users. Therefore, supplement the time-stamped test cases with at least two anticipated user roles. Both users work on different cells, followed by targeted work on the same critical cell. Document the visible state, centrally stored status, and behavior during concurrent calculations or validations. The expected conflict resolution approach shall be predefined.

In addition, verify changes to dependent inputs. Operator A enters a mass value, while Operator B modifies the dilution. A temporarily displayed result may be computationally accurate but still incomplete for decision-making purposes. The technical rule must define when a dataset is considered complete and when the reviewer is authorized to use it. A colored indicator of other user presence neither confirms completeness nor guarantees a stable final value.

Sources: [2]

Example: shared laboratory calculation

A team calculates the concentration of a test sample in a work spreadsheet. Employee A enters the initial weight and batch number; Employee B inputs the dilution factor. Following this, a reviewer evaluates the results and input values. In this example, the evaluation may only be used once both input fields have been completed and the reviewer has explicitly confirmed the designated reference point. This rule constitutes a standalone process decision, not a general Excel function.

The test begins with known reference values and independently calculated setpoints. Subsequently, the dilution is altered during the review. It is expected that the process will distinguish between the evaluation of the previous state and the subsequent change, and will not permit unobserved usage. The specific approach must demonstrate how this is achieved technically and organizationally. If the architecture does not support a sufficiently stable closure, the intended shared processing procedure for this decision must be adjusted.

Own work aid – Example: shared laboratory calculation
CaseTargeted actionEvaluation of the evidence
Separate inputs Two users modify different areas Complete shared dataset
Same cell Simultaneous differing values Identifiable and manageable conflict of interest
Review is ongoing Critical entry is modified The verified status remains unequivocal
Network interrupted A client is operating with delayed status update No undetected use as complete
Formula modified Operator attempts logic change Prevention or controlled detection
After release Value is being edited again Traceable modification and new assessment

Sources: [2] [3]

Test loss of network access and the handling of local copies

Joint processing depends on available services and functional synchronization. Test network loss during data entry, intermittent synchronization, and subsequent reconnection. The user must be able to determine whether their work progress is centrally available. Define how an ambiguous state is handled prior to other personnel accessing the data. A green indicator at an earlier time point does not serve as evidence of the current state.

Consider downloaded or email-sent copies. They may bear the same name yet differ from the controlled original. Determine which file is primary and how external copies are identified. Also assess access from Teams: what matters is the underlying library and permissions, not the visible interface. Avoid uncontrolled replacement processes in which an employee allegedly resolves a synchronization conflict by overwriting or renaming.

Sources: [1] [2]

Version history and audit trail serve various functions.

Version history may provide access to earlier files. It does not necessarily show each critical change, its reason or the decision it affected. Check what is available at the required level of detail. A saved copy of the entire file and a traceable history of cell changes provide different evidence. Some processes need additional controls or another system.

Test the retrieval of a historical event using the intended reviewer role. Can entries, calculations, and released status be identified as logically related? Are details regarding required storage duration retained? Also consider recovery procedures and permissions for deletion or version limitation. Microsoft recommends that version recovery be performed only after collaborative editing has been completed. Operational controls must prevent unintended loss of current, unsaved work during such recovery.

Sources: [3] [2] [1]

Anchor controls and responsibilities in operation

Name the responsible parties for workbook logic, Microsoft 365 configuration, and professional usage. Changes to formulas, macros, external links, or clients may trigger different evaluations. Also, library settings, sharing links, and retention rules are part of the operational concept. A regular review should consider actual usage: are shadow copies being created, do guests edit critical files, or is the intended closure being circumvented?

Train personnel on handling conflicts and ambiguous storage states using concrete examples. The work instruction must be actionable in daily operations. "Ensure synchronization before use" is too vague when no one can reliably identify a missing state. Define the verification steps and escalation procedure. A process that requires manual confirmation at every input—despite existing formula-based testing—may still be impractical in routine operations.

Sources: [2]

Make clear boundaries for qualification decisions

Summarize the intended use, supported environment, test results, and operating procedures. Explicitly identify any usage that is not covered. A jointly developed planning table may be appropriate, while the same architecture for a binding release calculation may require additional measures. This distinction is more helpful than the general statement "Excel in the cloud is validated."

If traceability or system control is technically insufficient, evaluate alternatives: separate data entry and analysis, controlled transfer into a central system, or a dedicated application. The decision shall be based on the process and risk assessment. Additional manual signatures do not automatically indicate an inadequate data architecture. Retain review evidence alongside the released status and establish triggers for a renewed evaluation.

Sources: [2]

Project experience and collaboration

Control simultaneous edits and identify the authoritative version

Co-authoring affects how a workbook is edited, saved and ultimately relied on. With IT and process owners, we trace simultaneous changes and test connection interruptions and conflicting local copies. A controlled location and clear editing rules may suffice for a limited collaboration. If concurrent changes or later reconstruction are difficult to control, we examine a different division of work with you. The decision should reflect the calculation and the consequences of an error.

Frequent Questions on Joint Excel Editing

Is AutoSave a release justification?

No. AutoSave saves changes during editing but does not confirm completeness or technical accuracy. The process requires a clearly defined review and closure stage, along with rules for subsequent modifications. These controls are verified against the actually used work folder and environment.

Sources: [1]

Does the Excel version history suffice as an audit trail?

A file version alone does not establish the required change history at the data level. Verify identity, timestamp, and reason for a critical change; EU GMP Annex 11, Section 9 expects documented justification for changes or deletions of GMP-relevant data. Compare the required information against the actually available version data. Remaining gaps require an appropriate solution and a justified determination of suitability.

Sources: [3]

Must every supported Excel client be validated?

Assess the actually authorized and relevant clients. If desktop and browser environments exhibit different functionalities or failure modes, these differences must be evaluated. Clearly define usage boundaries when only one client is covered. A non-enforceable restriction in a SOP is insufficient.

Can results be released during concurrent editing?

Only when the process reliably determines the authoritative status and protects against undetected changes. Review concurrent entries and modifications during the review process. If the assessed status remains unclear, the procedure must be adjusted prior to its use for the relevant GxP decision.

What precautions should be taken for guests and external personnel?

Verify identity, access rights, required training, and visibility of changes. Approval links and guest accounts require controlled issuance and revocation. The assessment also considers whether local copies are generated and which information must remain traceable after contract termination.

Sources: [2]

Must co-authoring always be disabled?

No. The key factor is suitability for the specific process. Joint handling may be appropriate when states, conflicts, traceability, and closure are well managed. For critical records, a different data capture or release architecture may be more suitable. This decision is risk-based.

Sources and review date

As of October 14, 2026. Regulatory classification and manufacturer responsibilities are separately documented. Examples, test matrices, and templates are standalone reference tools; they are customized for actual usage purposes.

  1. Microsoft: Collaborate on Excel workbooks with co-authoringSupported clients, cloud storage, AutoSave, and version recovery. Verify features according to the current Microsoft 365 environment. Checked on 14.10.2026.
  2. European Commission: EU GMP Annex 11, Computerised Systems (2011)Current version as per EudraLex inventory as of 14.10.2026. Application, risks, operation, and data lifecycle; specific testing proposals are derived independently.
  3. PIC/S: PI 041-1, Good Practices for Data Management and Integrity (2021)Inspection Guidance on Data Integrity; in particular electronic data, transmission, access, and review.

Sources checked: .

Daniel Herrmann, Daniel Herrmann Consulting

Daniel Herrmann Consulting

How we can support you

Daniel Herrmann Consulting combines professional assessment with hands-on work on the specific system. During the first meeting, we clarify your intended use, existing documentation, and the next most appropriate action step.

  • Computer system validation →

    Bring your GxP systems into validated operation, from laboratory software, MES and eQMS to ERP. We define the scope, plan risk-based tests using GAMP 5 and hand the evidence to QA for release.

Send a project enquiry Customer testimonials →