Introduction: Why FP&A Needs to Prioritise Data Quality

Analytics and AI are essential in today’s data-driven FP&A environment. These analytics and AI solutions depend on quality data. Yet, many FP&A professionals spend more time cleaning data than analysing the insights from analytics and AI. Why? Because the data is often incomplete, inconsistent, or unreliable.

Over 80% of work in analytics and AI is spent on data preparation. But what exactly is Data Preparation? Data preparation involves cleaning, transforming, and organising raw data into a format suitable for reporting, analytics, and AI. But preparation alone isn’t enough. The data must also be validated for quality to ensure it is fit for purpose.

Introducing the CACTUS Data Quality Framework

Data quality refers to the fitness of data for its intended purpose, i.e., operations, compliance, and decision-making (with reporting, analytics, and AI). Given the enormous amount of data firms originate and capture these days, how can the quality of data be improved? Given that assessment is the first step in improving the quality of data, to assess and improve data quality in FP&A, the CACTUS framework offers six practical dimensions:

C – Completeness: Are all required values present?

A – Accuracy: Is the data correct and verifiable?

C – Consistency: Is data coherent across time, systems, and teams?

T – Timeliness: Is the data up to date when you need it?

U – Uniqueness: Are there duplicates confusing?

S – Stability: Is the data structurally and logically reliable?

Figure 1 below summarises these six dimensions. The visual emphasises that all six must be considered together, as weaknesses in even one area can compromise the reliability of FP&A analytics. 

Figure 1. CACTUS Data Quality Dimensions 

Why does CACTUS matter for FP&A? Analytics and AI cannot run on broken data. Without trust in the data, any analysis is just noise. CACTUS doesn’t just help you identify problems. It guides how to fix them, ensuring that your data foundation is solid before it is used in analytics or AI models. Basically, these six dimensions provide a structured way to evaluate whether the data is “business-ready” for analytics and AI.

Real-World Example: CACTUS in Action

Let us take the case of an FP&A team preparing a “Budget vs. Actuals” report for the quarter. The reliability of the analysis depends on the quality of the underlying data. This is where the CACTUS framework comes in:

Completeness:
Are all the required data elements available? For example, do all General Ledger (GL) accounts, cost centres, and forecast line items have entries? Missing revenue or expense data from a business unit could distort variances and mislead decision-makers.

Accuracy:
Are financial policies correctly applied? This includes ensuring exchange rates are aligned with treasury guidelines, allocations are performed as per corporate policy, and no manual overrides distort the numbers. If exchange rates are applied inconsistently, international markets' actual vs. budget variance may look inflated or understated.

Consistency:
Is terminology and coding uniform across systems? For instance, if “Marketing Expense” is labelled differently in ERP and the CRM tools, reports may fragment and confuse stakeholders. Standardising naming ensures apples-to-apples comparisons.

Timeliness:
Is the actual data current and refreshed on schedule? If the FP&A team is working with data that is three weeks old, the analysis will miss recent expenses, accruals, or revenue adjustments, undermining its relevance for management reviews.

Uniqueness:
Are there any duplicate records? For example, if a journal entry is posted twice in the actuals feed, the expense line item will appear inflated. Deduplication safeguards against false variances and wrong conclusions.

Stability:
Has the underlying data structure remained unchanged? If the ERP vendor changes the chart of accounts format, historical trend comparisons may break, requiring re-mapping. Stable structures are essential for maintaining continuity in long-term variance analysis.

By systematically applying CACTUS, the FP&A team ensures that the “Budget vs. Actuals” report is reliable, timely, and trusted. This delivers three tangible benefits:

1. Improved Operations. Leaders can act quickly on true variances (e.g., overspending in a department) rather than chasing data errors.

2. Better Compliance. Financial policies, allocations, and reporting standards are consistently applied, reducing audit risks.

3. Stronger Insights for Analytics and AI. High-quality, trusted data feeds advanced forecasting models, scenario planning, and predictive analytics, making the insights credible and actionable.

In short, CACTUS acts as a quality shield, ensuring that finance teams base their analysis on quality data.

Conclusion

In FP&A, delivering faster, smarter insights means starting with high-quality data. The CACTUS framework equips FP&A professionals with a structured approach to assess and improve data quality across systems and time. Without it, even the best analytics tools or AI models won’t perform as expected. 

So, once you have ensured data quality, how do you make that data usable, shareable, and accessible across teams? That’s where the FAIR Data Management Framework comes in. The next article, “Make Your Data Work for You — The FAIR Framework for FP&A”, looks at making quality data ready for consumption or utilisation.

About the Author

Dr. Prashanth Southekal is the author of three acclaimed books — Data for Business Performance, Analytics Best Practices, and Data Quality. His second book, Analytics Best Practices, was recognized by BookAuthority in May 2022 as the #1 analytics book of all time, reflecting his global thought leadership in data and analytics.