The Value of a ‘Forecasting First’ Approach
James Kipling is Product Manager at Quantrix, a cutting-edge technology provider producing the next-generation spreadsheet. Here, James outlines the importance of good forecasting and gives an insight into its impact on businesses. The rise of ‘Big Data’ has seen technology companies, of all shapes and sizes, vying to increase the speed and effectiveness at which […]
James Kipling is Product Manager at Quantrix, a cutting-edge technology provider producing the next-generation spreadsheet. Here, James outlines the importance of good forecasting and gives an insight into its impact on businesses.
The rise of ‘Big Data’ has seen technology companies, of all shapes and sizes, vying to increase the speed and effectiveness at which they analyse trends in customer data, production data and macro-economic financial data. In order to identify trends and areas where they can gain efficiencies, these companies are likely to analyse all of the data sources available. So, what’s the problem?
Put simply, analysing data and trends with business intelligence tools alone does little to highlight the opportunities available for a business to capitalise on – meaning time, money and resources are frequently wasted. In a competitive environment, it’s easy to be left behind. But, there is a way forward – through superior business forecasting.
There are many types of business forecasting – demand planning, sales forecasting, inventory planning, capacity planning, and financial forecasting, to name just a few. For many companies, forecasts guide and drive business activity, so it is vitally important forecast models are based on reliable data, cover the full spectrum of likely scenarios and are integrated with the rest of the business to show the full ‘cause and effect’ of scenario changes as they ripple through an organisation.
It’s not about having all the data, but having the relevant data.
The most successful companies facilitate the bi-directional flow of information between business intelligence (looking at historic data) and forecasting (modelling the future) functions within their organisation. Beginning with robust basic forecasts to test assumptions, an opportunity is identified, forecasts are refined with the aid of historical data and trends, action is made to capitalise on the opportunity, and critically, the results of the action are fed back into the future forecasts to further refine their accuracy and effectiveness.
Too often, companies do not begin with accurate forecasting and the effects can have dramatic and far reaching consequences.
“There are so many challenges, really, and people tend to make emotional rather than objective decisions,” says George Pappas, a venture capital-affiliated software executive. “They also tend to make their models match their expectations. But really, the levers that drive results are too complicated in most cases to model in a program like Excel.”
Take the real-world scenario below of a company developing assumptions for new customer growth, including the sales cycle, close rate, deployment time, and economics (Table 1).
Through a flattened Excel view of the sales booking of these new customers, the assumptions seem reasonable (Table 2).
But then you look at result of these assumptions. A multi-dimensional model shows the reality: if you sell this way and your deployment time is as predicted, you must be prepared to handle rolling out to 70 locations in one month at peak load and staffing.
This one example clearly shows the benefits of good forecasting. Without this insight, the company profiled – and still in its early days of development – would have been seriously hindered.
Given the complexity of today’s businesses, companies should be wary of the tools they select for forecasting. Indeed, many are using solutions that simply aren’t sophisticated enough to handle the task at hand.
The go-to tool for forecasting in almost all industries is the traditional spreadsheet, meaning businesses open themselves to well-documented risks such as calculation performance, a lack of transparency and perhaps the most pressing issue of all – errors. And then, once a company has invested time to build a spreadsheet model robust enough to capture the key drivers of the business, often the nature of the forecast has changed.
Forecasting is time consuming, and can involve consolidating, summarizing, communicating, explaining and reviewing. Unfortunately for businesses, the hours and weeks spent creating forecasts means they can potentially become redundant after completion – with forecast-horizons passing or the initial conditions changing. Because of these challenges, the frequency of accurate forecasting within a company suffers.
Some successful companies incorporate a minimum of an 18-month rolling forecast. This allows them to ‘peek’ around the corner early on, leading to an increased speed of planning and budgeting – generally these companies are well-oiled ships and start by forecasting the future as a basis for all decisions. But such companies are rare.
Only 19% of survey respondents to the PwC budgeting and forecasting survey used a “best-in-breed” planning and forecasting application to aid in the rapid creation of accurate forecasts. In the same survey, ATK’s director of finance, Michael Varecka, draws interesting insights into the mindset of analysts, suggesting one way to ease the transition away from spreadsheets is to introduce best-in-breed systems which can accommodate a degree of personalisation whilst staying close to the financial truth. The goal is to reduce spreadsheet reliance, not to fully eliminate it.
The message to businesses needs to be loud and clear. If organisations lack confidence or accuracy in their forecasts, their conservative approach will lead to missed opportunities – or on the flip side – an overly optimistic forecast will mean they rely on opportunities that simply don’t exist. But get forecasting right and you’re in for a much easier ride.