Data Quality in Financial Reporting: Why Copilot Alone Is Not Enough

The data quality in financial reporting is crucial. The figures in the report must be traceable and correct. With the rise of AI assistants, the question arises as to how these new tools will influence the work of the finance team. Tools like Microsoft Copilot promise efficiency for individual tasks. However, they do not automate the entire process. Securing data quality across the entire chain — from the ERP export to the finished report — requires a structured approach.

The Role of AI Assistants in Day-to-Day Finance

AI assistants are designed to perform specific, defined tasks. Microsoft is integrating this technology to reduce the fragmentation of workflows and accelerate decisions (Microsoft 2026). A controller, for example, can use an assistant to draft an email or suggest a complex Excel formula. The interaction is based on a specific instruction, a 'prompt'.

However, the responsibility for the result remains with the user. The assistant only understands the context to the extent that it is communicated. It does not check the origin of the data or validate the business logic behind it. If a response does not meet expectations, the user can have a new response generated (Microsoft 2026). This shows that humans must actively manage and check the results. An AI assistant is a tool for a task, not a guarantor of a correct process.

Process Integrity Versus Point Automation

Financial reporting is a process. It is a chain of work steps that build precisely on one another. The chain begins with the data export from the ERP system, be it SAP, Abacus or Dynamics. It leads through various stages of data preparation, consolidation and analysis in Excel. At the end are the finished reports for management or the quarterly financial statements.

The integrity of this chain is the basis for reliable figures. An error in one place propagates and can distort the entire result. This is where the difference between point assistance and process automation lies. An AI assistant can speed up a single step, for example, formatting a table. The transitions between steps — copying data, manually adjusting references — remain prone to error.

End-to-end process automation, on the other hand, considers the entire chain. It ensures that data flows from one step to the next without manual intervention. The process steps are documented and repeatable. The logic is fixed and is not reinterpreted with each execution. This creates a traceable and more auditable environment for reporting. Quality is embedded in the process and is less dependent on an employee's form on the day or the interpretation of an AI tool.

Data Quality in Uncertain Times

In an uncertain economic environment, the value of reliable financial information increases. Decisions must be made on a solid basis. A recent study confirms that digital transformation has a positive and significant impact on the quality of accounting information (doi.org 2024). Structured digitalisation improves transparency, reduces information asymmetries and limits opportunistic management behaviour (doi.org 2024).

The study also reveals an important correlation. The positive effect of digitalisation on data quality is further amplified by uncertain environmental conditions (doi.org 2024). This means: the more unstable the market, the more important robust digital processes become for maintaining information quality. Companies that systematically digitalise their financial processes are more resilient.

However, this 'digital transformation' means more than just using individual AI tools. It refers to the structured redesign of processes through technology. It is about replacing the manual Excel layer that surrounds many ERP systems with a controlled, automated workflow. This creates the reliability that is required in uncertain times.

What Automation Does Not Do

Process automation for reporting does not replace the ERP system. The ERP should be the single source of truth for the accounting data. Skopa automates the Excel layer that exists around the ERP — the path from export via recurring reports to submissions. The ERP itself is not changed.

Automation also does not replace the judgement of the finance expert. A controller or CFO must interpret the figures, understand the story behind them and derive strategic recommendations. Automation ensures that the figures on which these judgements are based are reliable. The time previously spent on manual data preparation is now available for analysis.

Finally, automation does not eliminate the need for controls. On the contrary, it makes controls more effective. As the process is standardised and documented, control points can be set specifically and also automated. Auditing becomes simpler and more reliable because the steps in the process are traceable.

Concrete Steps for Your Reporting

You can improve the quality of your reporting this month. No external help is necessary for this. The following steps will help you to identify risks and increase control.

• Document the process: Map out the entire path from the ERP export to the finished report. Note down every single step, every file used and every person responsible. Be as detailed as possible.
• Identify manual breaks: Look for the points in your process map where data is manually copied from one file to another. Where are formulas adjusted by hand? Where do calculations take place outside of fixed templates? These are the biggest risk factors for errors.
• Assess controls: Review how and where the figures are currently checked. Are there documented plausibility checks? Are reconciliations carried out regularly and according to a fixed scheme? Often, controls are informal and not consistent.
• Check standardisation: Before you think about complex tools, check the potential for standardisation. Can Excel templates be improved and locked to prevent manual changes? Can a simple checklist for the month-end closing ensure completeness?

This analysis creates a clear basis. It shows where the greatest risks to your data quality lie and where you can achieve a major impact with simple means.

Microsoft, Copilot, Excel, Power BI and Dynamics are registered trademarks of Microsoft Corporation. SAP is a registered trademark of SAP SE. Abacus is a registered trademark of Abacus Research AG. bexio is a registered trademark of bexio AG. Sage is a registered trademark of The Sage Group plc. Odoo is a registered trademark of Odoo S.A.