In the second part of the Cash Flow Forecasting: From Strategy To Implementation series, we will uncover the pains and gains of information collection. Having timely access to correct information is a key requirement to successfully built accurate forecasts. However, finding the right information at the right time is not an easy task.

It becomes even more challenging if you are working in an environment with multiple entities, multiple currencies, multiple ERP systems, etc. Centralization has been pushed forward to counter these issues. However, I have not yet met that company, that is 100% centralized. There is always at least one region that does not fit within the centralized scheme.

As every bit of information is of the utmost importance when talking about cash forecasting, the spread of information usually results in a bunch of treasurers running around the company.

There should be a way out of this thread mill and there sure is one: Prepare yourself; step by step, to make sure you have all information and inputs available before you start creating the actual cash forecast.

Step 1: A never-ending quest to find the right data

The first step must be - without a doubt - the data definition phase: an organization analyzes and defines which types of data they want to include in the forecast.

An obvious data source is ERP data. Including this data is a must, as this will tell you exactly what drives your business. ERP data has the power to provide you with full visibility on the order-to-cash, purchase-to-cash and forecast-to-fulfillment processes.

On top of these operational cash drivers, salaries, loans, orders, leases, etc., which are not always residing in the ERP systems, should be incorporated in the cash forecasting calculation as well. More on that later.

One step further: cash flow logics

This sounds easy, right? However, if you want to take this one step further, there are important choices to make. For example, how to incorporate real customer payment behavior? Everybody knows there is an obvious difference between theory and practice. Companies will differ in the degree of analytics they use for developing payment scenarios. Important here is – once again – setting the right assumptions from the start, and aligning on the expected forecast accuracy. There is no problem in using theoretical payment behavior, if you agree with all parties (including the business) what the impact will be on the actual forecast. Other so-called “cash flow forecasting logics” may need to be applied: reflecting your payment runs, applying future rebate programs… We’ll come back to this particular in more detail, when we have a good view on the historical cash flows (see below).

Step 2: where and how to collect the data?

Once you have a good idea about what data to include, another challenge emerges: where to find all this data & how to collect it?

Not all information will be available at your fingertips. Some data will be easily accessible, for other information you might end up running from one side to the other side in the organization. In most organizations, ERP systems or Treasury Mgmt Systems will already aggregate a lot of useful data in one place. On the other hand, some information might be hidden in an exploded excel sheet or, even worse, in the mind of one of your colleagues. Carefully mapping the needed information to the right source won’t be an easy exercise, but when it’s done well, it will lay the foundation for a repeatable cash forecasting process.

Step 3: Learn from your past

When you have finally managed to get all relevant data on your desk, there is one more important step you should do: gaining insights in the current cash flows. It is probably less straight forward then the previous two, but it will make a great difference in the accuracy of the forecast if it is executed well. Knowing your current cash position is one thing, but having a view on the historic trends and seasonality will prove to be powerful information when taken into account in the cash forecasting process. If you look back in time, you will discover peaks and lows, some of which having inexplicable causes at first glance. Diving into the specifics of these fluctuations will bring you an invaluable amount of information, which you can use to finetune your cash forecast. A useful technique is to build a tree-structure. By descending the tree it becomes clear which are the main drivers and which are the highest contributing cost categories of your cash arms. If you want to predict the future, you must know what is eating or feeding your cash today. These cash insights don’t come easy and many companies still struggle to create an accurate view of their day-to-day cash position. Books can be written about the ideal (automatic) cash categorization set-up. In terms of cash forecasting, this is an important prerequisite if you want to take into account your company’s history.

“If you want to predict the future, you must know what is eating or feeding your cash today”

I’m stating the obvious here, when saying that each of these three steps is a tremendous amount of work on its own. But they are an indispensable step up when it comes to cash forecasting.

As indicated above, we haven’t spoken yet about the real deal: the logics behind cash forecasting. Where does all the data come together and how do we make an accurate cash forecast from each element? Please stay put, as this is something I will be discussing in my next post.