By Steve Morlidge, Business Forecasting thought leader, author of "Future Ready: How to Master Business Forecasting" and "The Little Book of Beyond Budgeting"
In this article, Steve Morlidge, author of "Future Ready: How to Master Business Forecasting", argues that the quality of business forecasting – used to steer an organisation – is unacceptably poor. He goes on to present six simple principles that will help executives significantly improve the performance of their forecast processes. More reliable forecasts speed up decision making and so help make businesses more agile.
The recent economic crash has been badly damaged reputations as well as fortunes – no more so than those of economic ‘experts’ who have been roundly criticised for failing to forecast the catastrophe. However, failures in forecasting are not confined to large-scale economic systems. The forecasts used by business executives to steer their businesses have also proved highly fallible. ‘The financial crisis has obliterated corporate forecasts’ reports the CFO Magazine (Ryan, 2009); 70% of respondents to their recent survey said that they were unable to see more than one quarter ahead. However, the problem is not restricted to times of economic turmoil. Over the last four years, the 1300 companies quoted on the London Stock Exchange issued, on average, 400 profit warnings every year. On average, each one resulted in a loss of 10 to 20% of market capitalization (Bloom et al., 2009); some $200 million.
It is, therefore, no surprise that a survey of 540 senior executives recently conducted for KPMG (EIU, 2007) found that improving forecasting came at the top of the priority list for the next three years. ‘Ability to forecast results’ also comes at the top of the list of ‘Internal Concerns’ for CFO’s right across the globe (Karaian, 2009).
CFO’s are right to be concerned; business forecasting is riddled with bad practices. For example, most businesses for much of the year forecast no further than the financial year-end. As a result, there is little visibility of ‘the road ahead’. Forecasts are often too detailed and too late for managers to take action. Boardrooms resonate with acrimonious debate about what is the ‘right number’, yet many organisations have multiple forecast processes, each presenting competing views of the future, which are never reconciled. Obsession with accuracy may coexist with a culture where professionally prepared forecasts are arbitrarily adjusted on a routine basis. Leaders often give contradictory messages, such as ‘give me your best estimate of what you think will happen’ and ‘your forecast must come back to target’, leaving managers confused and disorientated. Manipulation of forecasts as part of a corporate political game is rife; numbers are frequently ‘sandbagged’ or ‘spun’ to create a favourable impression.
At the heart of the problems experienced with forecasting is a fundamental misconception: that forecasting is the same as prediction. The role of forecasting is to provide us with information about what might happen so that we can take action to avoid the forecast outcome if it is not what we want. If we do so, we invalidate the forecast. Forecasts are for helping you to steer to your destination; they do not prophesy your fate.
The failure to grasp the fundamental nature of forecasting is compounded by a second misconception. In my experience, managers either believe that forecasting is straightforward – ‘just common sense’ – or that it is extremely complicated - requiring the use of complex mathematics – and so best left to experts. The reality is that it is neither; it is a matter of properly understanding the nature of forecasting as an aid to decision making, and working in an organised and disciplined way to produce ‘good enough’ forecasts. Good tools and techniques may be necessary, but they are not sufficient. You need to know how to use them properly and create the kind of culture that encourages people to tell the truth.
Forecasting in business is a complex mess, but it need not be. I believe that there are six simple principles, which once mastered, will significantly improve the quality of forecasting in almost any organisation.
Business forecasting is like sailing at sea. It makes sense to plan before you start the journey, but the original plan is often soon out of date because of changes in the weather or tides. At this point, you need to forecast where you are headed, so that you can work out what corrective action is needed to get you to your destination.
The first thing that is clear from this example is that it is important to make a sharp distinction between a forecast (where you think you will be) and a target (where you would like to be). The second thing it helps us to understand is the role of forecasting: to support decision making. In order to do this well, a forecast needs the following qualities. It should be:
How far ahead do you need to forecast? The answer depends on how long it takes to enact a decision.
The captain of a super tanker needs to consistently forecast 3 miles ahead because that is how long it takes it to stop. A speedboat, on the other hand, may require very little forward visibility. In practice, this means that businesses need a rolling forecast horizon, based on the lead times associated with ‘steering actions’. A traditional year-end forecast is like overtaking on a blind bend – you have no idea of the possible outcome of your decision.
How frequently should you forecast? That depends on how quickly things change. More frequent forecasts are needed to safely navigate through the busy Singapore Straights than in the wide open seas of the South Pacific.
Any form of forecast requires a model; a set of assumptions about the way the world works. The model used in forecasting could be a statistical model; one that extrapolates into the future from the past. This approach can be effective, but often the future is not like the past. You might, therefore, choose to use a mathematical, or driver based, model; for example one that helps you to forecast the impact of volume on the cost base of the business. However, often the world is too complex, or the business too fast changing to make this approach workable. That is why forecasting in business often relies heavily on judgment; where the model is in the head of an expert or a larger number of people who ‘know what is going on’. This approach is not without problems. Human judgment can be flawed, and managers can feel under pressure to adjust forecasts to ‘avoid giving nasty surprises’ or ’sounding defeatist’. As a result, judgmental forecasts are particularly prone to bias.
The trick is to understand the range of methodologies available, choose the appropriate one, and take steps to mitigate its weaknesses. So, for example, a statistical or mathematical technique might be used to produce a baseline or ‘business as usual’ forecast and judgement to estimate the impact of the decision made to change the course of affairs.
The only guarantee that you can rely on a forecast to make decisions that affect the future is that those previous ones have proved to be reliable in the past. Yet, few businesses take the simple steps required to monitor their processes for evidence of bias, so that they can take action to eliminate it if detected. Most businesses fail to measure forecast quality at all.
Those businesses that do attempt to monitor forecast quality often measure the wrong things at the wrong time. A common mistake is to measure forecast errors at a point in time that is likely to be after the forecast has been acted upon. This is like blaming the navigator for having forecast a calamity that her forecast has helped avert. At its simplest, a series of four short-term errors with the same sign (positive or negative) is evidence of bias; any fewer is likely to be the result of chance.
The only thing that we know with absolute confidence about the future is that any forecast we make is likely to be wrong! Where there is the debate about the forecast, it should not focus on whether you have the right ‘single point forecast’, but how it might be wrong, why, and what to do about it. In particular, it is important to distinguish between ‘risk’ - random variation around a realistic single point forecast - and ‘uncertainty’ – resulting from a shift in the behaviour of a system that completely invalidates the forecast. Banks’ over-reliance on risk models that failed to take account of uncertainty was a major contributor to the recent economic collapse. Whatever form your ignorance of the future takes, it is important to develop the capability to spot and diagnose deviations from forecast quickly, and to create a ‘play book’ of potential actions to enable a swift and effective response.
Mastering forecasting is not an art, but neither is it complex science. It is mainly a matter of applying a modest amount of knowledge in a disciplined and organised fashion; as a process. A good process – like a good golf swing - will produce good results.
Building a good process involves doing the right things in the right order (cultivating a good technique), in the same way over and over again (grooving the swing). Those things that are responsible for bias (hooks and slices) should be designed out of the process (remodelling the swing), and the results of the process continuously monitored (the score) and minor flaws corrected as they become evident Again, like golf, temperament is as important as technique. Blaming people when the process is at fault is a sure way to encourage dishonest forecasting
Done well, forecasting will help a business respond swiftly and effectively to emerging reality and so gain a competitive edge. Done badly, management may be misled into making the wrong decisions. However, businesses have little option but to forecast, because without any kind of ability to anticipate, organizations can only react to those things that have already happened, which, by definition, they have no ability to influence. The tools and techniques that business need are already available, managers simply need to learn how to use them effectively.
Copyright, ©2018 fpa-trends.com. All rights reserved.