Sales forecasting is quite a debate on its own, and recently, business media drew a divide between people who forecast with dedicated software versus those who still use Excel for forecasts (if you haven’t seen this article, you are missing out!). So, we thought we are going to put our two pence in the debate about using Excel for forecasts and business forecasting.
Most sales forecasts require two sorts of estimations: adding sales and multiplying growth rates — or, on occasion, reductions. Sales are seldom steady, and many markets have fluctuations in sales throughout the year. So, it sounds like it is pretty simple data input job, with a couple of graphs from the first glance. However, some advisers and financial planners aiming to forecast in an uncertain or changing environment require quite modelling-intensive solutions and that is where the problems arise.
While Excel has been updated dramatically since it first hit the market in 1985, but it does still have its limitations. Many corporate accountants and financial planners said they remain big fans but admit to using specialised software for some tricky or more modelling-intensive jobs.
One of the principal issues of spreadsheets is that no all-inclusive system is available. This makes their utilization for fundamental applications – eg time and cost administration, client databases – something of an issue. Also, with regards to forecasting and any calculations, spreadsheets can be a minefield.
Additionally, using someone’s else’s spreadsheet is troublesome. This is particularly important when it comes to company-wide reporting, using the same spreadsheet through multiple teams is bound to creep up some problems and confusion among the teams. Furthermore, in such scenario trying to find information in a large model is difficult. Every reader will want to focus on different data and review a model in different ways. Unless you have the time to go through validating every formula and checksum, there is no real way of knowing a hand-built Excel model is accurate.
Although, few businesses are able to take concrete steps to ensure best practices for their Excel master sheets, when they don’t – problems arise. Unfortunately, it is generally accepted that nine out of every 10 spreadsheets suffer from some error, and when used for financial reporting consequences can be severe.
In the case of large corporates, they can afford large application where all of the above problems are solved by the use of multi-dimensional database tools. This provides the user with much greater insight and is driven by planning, budgeting and forecasting, which is integrated into wider corporate performance management (CPM) suites. However, smaller businesses often cannot allow themselves to have an in-house CPM suite that is often run on local networks. So smaller companies are left to rely on spreadsheets and inconsistent reporting throughout the year.
For modelling intensive solutions, a classic advice online is to choose Excel for forecasts that uses sheets and templates. But those tasks are often data heavy even for small businesses. Modelling-intensive forecasts are made based upon judgements, informed by an understanding of the market, competition and the dynamics of the business. Historic data is important in these situations, but more crucial is the requirement for the model to respond to changes in assumptions, and to be capable of being continually altered and improved. And having so much data in one or a couple Excel sheets leads to problems.
People tend to look away from software solutions, in modelling intensive cases, because they need the flexibility to adapt their forecasts. At ProForecast we offer a forecasting software, that doesn’t cost a fortune, and allow you to change only fractions of your data to adapt to any changes that business might require. Also, if the forecasting task is data-intensive, Excel forecast is not going to give you the required control and structure over a lot of data and should not be used. A robust software package is more suitable.
So, to answer the age-old question is time to move on from Excel for forecasts? Yes, but only if you run more advanced forecast that is data heavy and modelling intensive, rather than just simple sales projections.