3 Tips for Running Better Ad-Hoc Analytics on Business Financial Data
Performing ad hoc analysis is very important for modern business enterprises
As one of the best sources of data for business analysis, a company’s financial data forms the basis for working capital projections, required reporting to tax authorities, and even overall business strategy.
Indeed, financial data contains a wealth of information for companies looking to increase their footprint. A company’s financial condition provides FP&A teams with the ability to empower CFOs with the insights they need to build resilient businesses.
However, analyzing financial data is difficult, especially on an agile basis. As companies collect more data than ever before, gaps in data governance and data quality are hampering analysts’ ability to dig deeper into datasets. You can’t trust the results of your what-if scenario projections if you only have access to the last quarter’s metrics.
As the pace of business accelerates, senior executives need on-demand information to make informed decisions about market priorities and opportunities as they arise. The need for fast insights therefore sheds light on ad hoc reporting, where analysts must query and retrieve data on the fly.
When coupled with deficiencies in data analysis processes and lack of training in data science principles, most companies run the risk of drawing the wrong conclusions. Here are three tips for doing better ad hoc analysis of your financial data.
Review revenue data sources
Companies these days collect revenue from dozens of sources, making ad hoc analysis difficult. Revenue data comes in many formats and structures, and it’s difficult to ensure that this data conforms to your own storage schemes.
However, the problems with these datasets go far beyond storage issues. Often, data from revenue sources arrives in forms that do not lend themselves well to ad hoc analysis.
For example, revenue collected from iOS or Google’s Play Store comes as a lump sum of data, broken down only by transaction without further context. Examining the context of the data set forces you to dig deeper into application interaction metrics and customer data repositories if you want to correlate them effectively with revenue trends. In contrast, POS revenue data tends to be detailed and contain high granularity.
Standardizing the granularity of revenue data is essential if you want to run analyzes of all data sources. If you don’t, you’ll create data silos that provide incomplete views of your revenue. Worse still, you will have to manually transform and transfer data from one source to another before running analyses.
These processes introduce analysis errors that lead to misunderstanding. These tedious processes only add friction to ad hoc analyses. For example, you can’t run ad hoc reports on datasets stored in spreadsheets, because the rigid nature of a spreadsheet doesn’t allow you to change parameters on the fly or easily dig into datasets stored elsewhere with new queries.
You also can’t integrate real-time revenue data into your reports, since each silo will be updated at different times? Standardizing the granularity of revenue data may result in loss of information on some sources. However, it will allow you to automate data collection and cleaning, allowing you to have flexible ad-hoc reporting functionality.
Use outcome measures
Analysts often struggle to convey the impact of their models on the business. A big reason for this is the use of irrelevant metrics. For example, using a non-quantifiable metric such as customer satisfaction to convey the impact of an innovative financial model is pointless in an ad hoc report because senior executives cannot directly link customer satisfaction to sales and revenue.
Outcome metrics such as ROI and IRR cut through the fog and speak directly to the performance of a company’s investments and projects.
For example, comparing YOY revenue growth is the norm at most companies. However, if your business is a high-growth startup, it makes more sense to compare quarterly revenue instead of yearly numbers in your ad-hoc reports. Some start-ups will benefit from comparing monthly revenue because the growth they experience is exponential.
While ad hoc reporting gives you the freedom to explore data on the fly, it’s important to define the metrics you’ll use up front to ensure you’re always measuring the right results. In short, explore your data, but be careful when adopting metrics or benchmarks as you go.
When using these metrics, make sure you understand what they convey. For example, ROI and IRR quantify returns, but they measure completely different conditions. ROI measures the overall return, while IRR measures the equivalent discount rate in a NPV calculation. In ad hoc scenarios, ROI can provide better insight into the longer-term goal of an IRR. The IRR requires time and discount rate inputs to give it more context.
In ad hoc scenarios, the hasty estimation of these figures can generate errors in the calculations which exaggerate the results. The discount rate you arrive at (the goal of an IRR calculation) can be wildly off track over time. Macro factors such as central bank interest rates can make your calculations stale. A simpler return on investment calculation will provide more flexibility and a quick overview of the attractiveness of an investment.
Quantify all data points using dollar amounts for maximum impact in your reports and use data visualizations to communicate results when reporting to senior executives. Although the financial world relies on data tables, they do not lend themselves well to quantifying business impact on non-financial audiences.
Consider revenue model biases
When running ad hoc reports, pay attention to the impact of bias on the final results. By definition, ad hoc reports look at business impact right now and use real-time data. However, the diversity of a company’s revenue models can significantly skew the results.
For example, analysts at a company that relies on SaaS revenue will typically witness consistent trends for the most part. You’ll see costs spread relatively evenly, while revenue will also be predictable, assuming the company’s products see traction and the acquisition funnel is buzzing.
However, a freemium model works differently. Revenue per user will increase significantly as users upgrade to premium features after the free trial period in cohorts. Costs per user are also higher initially in this model because the business will collect less revenue, leading to skewed margins per user.
So, when designing ad hoc reports, you need to decide whether it makes sense to measure business impact per user, given the volatility of user counts.
Keep the big picture in mind by considering seasonal trends and pricing model adjustments when designing ad-hoc reports. After all, a company’s revenue model can also create trends that impact consumer behavior. So, always consider the context in which revenue and financial data is collected.
Challenging but rewarding
Financial analysis is difficult, but the results of ad hoc financial reports help a company capture a snapshot of its performance. Remember to review biases in the data and install appropriate governance processes before drawing conclusions from ad hoc reports. Follow the tips in this article and you’ll get in-depth insights that will future-proof your business.