So, it’s January 2025, and you’ve embraced the ‘new year, new you’ approach. You’ve rejoined the gym, taken up yoga, and sworn off chocolates until Easter. You’re also back at work, and your business is kicking off a Finance transformation programme with the same enthusiasm.

You’re planning to transform your systems, streamline processes, upskill your team, and dive into advanced analytics and artificial intelligence (AI). But here’s the thing: the success of any Finance transformation hinges on one often-overlooked factor—the quality of your financial data.

Whether you’re working for a global blue-chip implementing a Tier 1 ERP, or are an SME operating in a niche market implementing a Tier 3 solution, you’re going to need to think about your data from day 1…

At the heart of every Finance system lies a foundation of data structures and hierarchies. These are what define how financial information is categorized, aggregated, and reported. Without well-thought-out structures, financial data becomes a chaotic, tangled mess that’s hard to navigate and interpret. Properly designed data hierarchies ensure consistency, comparability, and accuracy across financial statements and reports.

For example, a poorly constructed chart of accounts can lead to inefficiencies in reporting, multiple sources of the truth, and discrepancies during audits leading to costly overruns. Conversely, a solution with standardised data running like a golden thread running through application simplifies architecture by consolidating all financial data into one unified framework. This eliminates redundant data silos and ensures seamless consolidation of information across departments, regions, or the entire organisation.

Gathering all reporting requirements upfront is an essential first step. The Finance function’s primary role is to produce insights and reports, and understanding these needs from the outset ensures the solution is designed to deliver on them. By building in data design principles, standardisation, and minimizing the number of interfaces and systems, organisations can create a streamlined solution.

A well-designed system should eliminate the need for manual intervention and data manipulation, ensuring accuracy, efficiency, and reliability in reporting.

Picture redesigning your house and building a fantastic new extension. You’ve got planning permission, architectural drawings, and even interior design plans featuring Pantone’s 2025 colours of the year. But none of it matters if the foundations aren’t solid. This is exactly what happens when organisations attempt to implement advanced reporting and analytics on poor-quality data. Clean, complete, and standardised data is the bedrock of insightful and actionable financial reporting.

Data cleansing involves identifying and fixing errors, while deduplication ensures that redundant entries don’t skew results. Standardisation aligns data formats and definitions across systems, creating consistency. Without these steps, the reports generated by your shiny new financial tools will be riddled with inaccuracies, leading to poor decision-making and a lack of trust in your systems.

Making sure your data is entered correctly from the start is equally important. Preventative data entry controls like validation rules, drop-down menus, auto-complete fields, and mandatory fields can keep your data neat and tidy by stopping your team from inputting information in the wrong format, making sure all critical fields are left completed. Most modern ERP systems have inbuilt functionality for real-time validation, and audit trails that track changes can further ensure accuracy. Directive controls like training your team on proper data entry protocols and your data policies is crucial, as is implementing detective control mechanisms like reviews of critical dataset entry and automated notifications for incomplete entries. Designing these data controls into your transformation design principles will go a long way towards minimising the risk of inaccuracies that compromise the integrity of your financial data.

Too often, Finance teams overlook the importance of data during the early phases of a transformation programme. This oversight can lead to significant delays and degraded outcomes as data issues surface mid-implementation. Fixing these problems mid-implementation will need rework and additional resources, detracting from the program's overall efficiency and success. Fixing these problems post-implementation often means embarking on the dreaded remediation programme to unpick the issues and redo the work - painful, costly, and entirely avoidable!

Don’t define how and what you want to be able to report out of your shiny new Finance function? You end up with a fat General Ledger, using internal orders to create reporting, duplicate master data records, and if you’re running inter-company and consolidation processes, they end up as a great big can of worms. Starting with a clear focus on data - including its quality, structure, and governance - ensures that your transformation program can avoid these scenarios, and creates a solid, stable, and sustainable foundation, allowing you to leverage all the fancy automation and AI in future.

Automation and AI promise to revolutionise Finance by improving efficiency and unlocking new insights. You can’t scroll your LinkedIn feed without coming across at least a handful of posts promising to free up your Finance team from repetitive, rule-based tasks to focus on higher-value activities. But these technologies are only as good as the data they ingest. Inaccurate, incomplete, or inconsistent data can cause automation scripts to fail and AI models to produce misleading or downright incorrect predictions.

Common Finance RPA solutions include OpenText's Vendor Invoice Management, which automates the matching of invoices with purchase orders and receipts to streamline your AP function, and BlackLine for account reconciliations, which eliminates tedious manual processes. Tools like Dext or SAP’s Concur automate expense processing, while Kyriba's treasury automation tool streamlines cash flow forecasting and liquidity management. Tier 1 ERP systems leverage algorithms to automate high-value tasks; Oracle’s Fusion Financials uses algorithms to automate daily cash forecasts, and SAP’s Financial Close automates close activities, speeding up your close timetable and freeing up your team to focus more on the judgemental elements of your Record2Report process. QuickBooks and Xero accounting software both allow integration with various RPA applications, as well as their own built in automation functionality.

However, poor data quality can cripple these solutions. Inconsistent vendor data can cause invoice matching tools to misidentify entries, while incomplete records in reconciliation systems lead to inaccurate financial close processes. AI-driven forecasting tools rely on clean historical data to identify patterns and predict future trends; flawed inputs mean unreliable forecast outputs.

Effective RPA and AI solutions rely on data with certain key characteristics to perform well. Think back to your ICAEW exam days, and the tests around account balances - then apply that to your Finance transformation programme data. Good data must be:

• Complete: containing all relevant information necessary for accurate decision-making
• Accurate: free from errors or omissions that could send your outcomes awry
• Representative: reflecting the real-world conditions in which your AI or RPA system operates
• Objective: free from bias or discrimination that could skew results.

Meeting these criteria gives you the best chance that your automated processes and intelligent systems function as intended and give you the reliable insights and efficiencies that you have in your Finance Transformation business case!

While this article focuses on data, it would be remiss not to mention the ethical considerations surrounding AI and RPA in your Finance transformation programme. Solutions must be designed to avoid embedding or amplifying bias, as this can lead to unfair outcomes or reputational damage. Incorrect decision-making due to flawed data can result in financial loss or regulatory penalties. And let’s not forget data security - your financial data must be protected against breaches to maintain confidentiality and trust. Incorporating robust ethical guidelines alongside technical solutions ensures that these advanced tools operate responsibly and transparently.

Even the best data cleansing efforts at the start of your programme can be undone without robust governance frameworks. Data governance ensures your data remains accurate, consistent, and secure over time. It establishes clear ownership, policies, and procedures for maintaining data quality and prevents degradation due to mismanagement.

Key components of a governance framework include:

• Data ownership: Assigning responsibility for datasets to Finance, not IT. Finance teams understand the processes, reporting needs, and compliance requirements for their data.
• Data policies: Defining how data is handled and updated, including access controls to prevent unauthorized changes to critical master data.
• Regular audits: Tools like Informatica's Data Quality, Talend's Data Fabric, and SAS's Data Quality can automate data quality checks and identify discrepancies before they impact reporting or RPA routines.
• Training: Educating teams on the importance of data governance to ensure long-term adherence to best practices. Your Finance team may be experts on the intricacies of IFRS, but do they also understand the importance of data governance?

Embed data governance into your Finance Transformation programme to create a sustainable environment for high-quality data.

Finance Transformation is more than adopting new technologies; it’s about empowering your organisation to make better decisions, faster.

Defining your reporting requirements upfront is crucial - it means your system data can be designed to run like a golden thread, through all your applications, for all your outputs. You can eliminate redundancy, streamline reporting, and generate analytics to your heart's content that will be accurate, efficient, and reliable.

By investing in clean, standardised data and robust governance, you create the foundation for insightful reporting, effective automation, and reliable AI. Build in workstreams or work packages that define your data structure and quality upfront to save you time, effort, and frustration downstream. There are plenty of opportunities for frustration in a transformation programme - your data doesn't need to be one of them!

Your data isn't only an asset - it's the very foundation on which your programme is built. Take care of your data, and it will take care of your transformation!

Interested in exploring how to successfully deliver a system transformation further? Sign up to this webinar taking place in February 2025 to learn how to navigate core system change.