Service Insight: Producing a better sales forecast
pharmafile | March 21, 2011 | Feature | Sales and Marketing |Â Â Dr Hazel Rudge-Pickard, Pickard-Smith Consulting, Service Insight, forecastingÂ
SERVICE INSIGHT
Supply chains continue to be stretched as more and more of them go ‘global’. Combine these inherent pressures of language and systems differences with the current economic climate and it’s evident that the entire process needs a re-think for us to aspire to optimal performance.
Competition, choice and the volatility of customer demand seems to be a continual problem for most of us working in any supply chain, and it is becoming more difficult to address as expectations for lower costs, higher quality and improved responsiveness continue to grow. All the while, economic forces are encouraging significant reductions in end-to-end supply chain costs with rigorous inventory optimisation.
While forecast volatility and value certainly need some attention, arguably the greatest challenge currently facing the supply chain is to make the process and information visible to all concerned.
Although technology has moved on rapidly, few enterprises have the same structure that they were designed to service a decade ago.
Gone are the days when different departments could operate almost independently within their own silos. The demands placed on the modern supply chain mean that it has to cut right across departmental and functional boundaries and be embraced as a single process, clearly visible to the whole enterprise. Unfortunately, in many cases these new demands on process have left behind a collection of disparate tools, processes and systems that work poorly together and, in many cases, actually act to restrict visibility.
Why forecast?
Supply chain visibility is important because without it the wrong information is used by the wrong people at the wrong time; producing inaccurate forecasts and leading to a variety of problems like over – or under-stocking, missed deliveries, lost customers and ultimately, shrinking margins.
With closer monitoring of latest sales data and future demand, responses to market fluctuations could be much quicker and in some cases even anticipated. The main challenge has to be to improve the quality of the sales forecast.
Improving the forecast
In the late 90s when algorithmic forecasting from historic data became so popular, there were many who declared that the planning process had moved from an art to a science.
They thought that if we could develop sufficiently intelligent algorithms it would be possible to predict the future better than any human ever could. Quite soon it became apparent that whilst this approach worked well for some products, for the rest a ‘statistical’ forecast was wildly inaccurate.
The problem wasn’t with the tools, it was just that the tools didn’t have the full picture available to the expert planner.
Bringing together the right people, the right tools and the right data to create an accurate forecast is often very difficult. Sales forecasting is a task usually assigned to the people ‘at the coal face’ in the medium-sized enterprise – the sales team.
Forecasting is usually done on spreadsheets in the office once the latest sales figures have been released. And it is a job that many detest, feeling that time spent forecasting is time lost selling. Giving the forecasters tools that make the process quick and easy, is a vital step in helping them to produce a more accurate view on the future.
The second common barrier to retrieving a regular update on future sales demand is simply getting the forecasters into the office to make the updates. Modern, internet-based systems overcome this by allowing the forecasters to dial in from any location to adjust their numbers in minutes, letting them get back out on the road. By equipping them with tools to update their forecasts using apps on smart phones and tablets would improve matters even more.
Current wisdom suggests that a system generated baseline forecast from which a forecaster can work is the most effective and efficient way to work. Forecasting algorithms excel at picking up both global and seasonal historic trends and, by extrapolating them into the future, a reasonable baseline can often be achieved but it is the sales team that invariably has that deeper insight into their market. By baselining first, the task of the forecaster changes from one of drudgery, typing in potentially thousands of numbers, to expertly tweaking only those figures that experience will tell them are wrong.
Finally, and perhaps most importantly, forecasts should be tracked over time and measured for accuracy against actual data.
A forecast that is accurate for a number of consecutive periods allows us to reduce holding stock. Since demand patterns can be anticipated, production can be finely tuned.
A post-mortem of forecast accuracy each month during the Sales and Operations Planning (S&OP) meeting may show significant indicators in changing markets or unsatisfactory performance in a particular area of business. It can even be used to feed back to individual forecasters helping them to identify and correct problems with their forecasting technique.
A very common issue is that an individual may simply be too pessimistic with his forecasts. By identifying the problem through forecast accuracy analysis and showing him the evidence will often resolve the resultant under stocking issue down the supply chain line. It is also not uncommon for forecast accuracy to feed back into the reward scheme with the most accurate forecasters being moved onto the most important sales lines.
The bigger picture
The sales forecast should not be viewed in isolation; it should be seen as a starting point for an overall consensus plan – a single plan that is agreed by sales, marketing, product development, planning and production managers – although reaching a consensus can often be an interesting dilemma due to the often conflicting interests of each group.
Once a consensus planning system is in place it can then be taken further to provide networked S&OP where the forecasting and planning process is taken outside of the enterprise, to include customers at one end and suppliers at the other. In this case forecasts are shared with important suppliers and service providers to ensure that they will be able to meet future demand and provide information on potential delivery problems.
Customers are encourage to provide true demand driven S&OP and provide the supply chain with near real-time data to work with.
The goal of networked S&OP is still a long way off for many medium-sized companies.
Despite the rapid advances in computing, reliable and automatic data interchanges between disparate systems is still a problem and is likely to remain so for some time to come. The capture and analysis of data requires substantial effort and the sharing of that data or the results of any analysis can be fraught with difficulties.
Getting accurate and timely information flowing smoothly through the internal supply chain is a struggle for most, so the idea of extending that flow to encompass external suppliers and service providers causes sleepless nights. Larger enterprises have systems that consume and produce EDI data but tools understanding this format are generally out of the price range of more modest supply chains.
However, while some enterprises are still struggling to make different parts of their internal supply chain share information, the more advanced teams already have fully integrated systems where forecasts, plans and strategies are visible and acted upon right across the various business functions.
Market leading supply chains implemented by the most innovative enterprises have taken this a step further and have systems that use both internal and external data to provide ‘up to the minute’ key performance indicators (KPI) based on business intelligence.
Dashboards form a cornerstone of the next generation of supply chain systems, giving one-glance performance information to key personnel and management within the chain.
Automated performance management systems that constantly compare forecasts and plans against given business rules, take corrective action or raise alerts when conditions require. Automated exception notification provides the power of ‘management by exception’, relieving supply chain managers of the burden of studying reams of data for potential problems.
As supply chain systems become more intelligent it will be possible to extend exception notification further, so that warnings are given just after a problem has occurred but before such a scenario even arises. When a set of previously identified conditions that may cause a problem are detected, an alarm will be sent to the right person or an action taken automatically, without human intervention. This is when the supply chain moves away from being reactive to being proactive.
The future
Implementing a next generation supply chain system is difficult and will require significant investment of both time and money, but there are already strong indications that early adopters are reaping the rewards of their investment with increased growth, lower supply chain costs and greater competitiveness. Choosing the right tools to help achieve increased visibility and understanding how these tools integrate into the larger supply chain picture is central to the successful development of a next generation supply chain.
Dr Hazel Rudge-Pickard is service procurement consultant at Pickard-Smith Consulting. For more infomation visit: www.pickardsmith.com
For more information on Service Insight features contact InPharm’s sales team on +44 (0)1243 772 010 or email pharmafilesales@wiley.com
Related Content
Forecasting future launch success
Pharmaceutical companies are moving into new and sometimes very different (often specialised) therapy areas, because …
Service Insight: Logical logistics
SERVICE INSIGHTThere is no mistaking that we have been faced with turbulent times over the …
Service Insight: Submission development and market access
SERVICE INSIGHT A successful health technology assessment is critical for market access in the UK. …
