.svg)
<Read Part 1>
Rely on the data that matters
“In God we trust. All others must bring data.”
The famous quote from the modern-day polymath W. Edwards Deming refers to the need for decision-making within a business to be supported by evidence in the form of data. This is a well-known maxim within the world of FP&A but in the reality of the modern organization not all data is created equally, nor should it carry the same weight with decision-makers. This is due to several factors:
- The quality of capture for non-financial data can often be variable and is dependent on the training and scrutiny that is provided to the teams recording such data.
- Data provided by external providers can be absent of important context, information, or background which changes the value which should be assigned to either trends or outliers (for example a change in the methodology of data capture may shift the values created significantly).
- There is also a push within information technology to capture all data created within the organization, to create vast data lakes of structured and unstructured data.
Whilst this is useful for data scientists, this is not useful for FP&A teams looking to achieve high performance.
For FP&A teams, the data required is that which supports the logical business model of the organization. By creating and understanding this model (initially via the creation of value tree diagrams tracing the drivers of value throughout the business model) in conjunction with other teams within the business, the aim should be to identify the data linked to the drivers of the business which allows for operational insight to be derived.
Data external to this model should not be prioritised for inclusion unless it adds significant context to the operational data already identified. Much like how ruthless you need to be in becoming operationally focussed by removing the inefficient and the unused, high performance FP&A teams should make the best use of the data identified from the above process to remove as many as possible of the biases which enter into planning and analysis.
A practical example of this can be observed in revenue forecasting. Depending on the business, revenue forecasts can be either; a top-down activity driven by the CEO, CCO, COO or CRO (Chief Revenue Officer) using their experience within the business and understanding of the market to forecast revenue or a granular, bottom-up exercise with revenue forecasts consolidated from the sales representative upwards, massaged by the sales managers, tweaked by the head of a region and finally adjusted by the aforementioned C-suite person with responsibility for revenue.
In both cases, each forecast contains both the conscious and the subconscious bias of the people who have created it; what type of year are we having? How far away from the plan are we? Where does this put me versus the compensation plan? What does my manager/head of region/head of sales/fellow executives expect? Am I naturally cautious or optimistic? Is it nearing my annual review? The list is extensive but not exhaustive. As a result, the forecasts produced are liable to inaccuracies that cannot be explained by data held within the organization.
Alternatively, automated revenue forecasts based on CRM data (assuming this contains the leading indicators of business), combined with data from finance systems can remove these biases whilst also shortening the time taken to assemble a revenue forecast. However ultimate forecast accuracy is not the aim; improving performance is.
Blueprint action 2:
Base forecasts on data generated from the operational activities of the business (for example the sales pipeline), leadership and management teams are focussed on the drivers of performance within the business which can be improved via operational changes.
.png)
.svg)
