This site uses cookies to provide you with a more responsive and personalized service. By using this site you agree to our use of cookies. Please read our cookie notice for more information on the cookies we use and how to delete or block them.

Bookmark Email Print this page

At the Core of Forecasting Effectiveness

Improving accuracy in semiconductor forecasting

Because of the cyclical and volatile nature of the semiconductor industry, sales forecasting is perhaps one of the most critical factors for financial success year-in and year-out. An accurate forecast can help minimize the need to unexpectedly ramp-up or burn-off inventory, either of which can wreak havoc on profitability. Companies need one version of the truth to run their businesses from inventory to allocation of marketing funds to salesforce compensation. Learn how CFOs can look beyond the standard measures to focus on end-user demand forecasting, operational forecasting and long-term contracting to make sales forecasting more effective.

Meet our subject matter experts


Kim Frazier
Strategy Principal
Deloitte Consulting LLP



Srini Bangalore
Supply Chain and
Manufacturing Operations
Director
Deloitte Consulting LLP

Scott Angel
Semiconductor Practice Leader
Partner, Deloitte & Touche LLP


Transcript

Scott Angel: Because of the cyclical and volatile nature of the semiconductor industry, sales forecasting is perhaps one of the most critical factors for financial success year in and year out. An accurate forecast can help minimize the need to unexpectedly ramp up or burn off inventory, either of which can wreak havoc on profitability. Companies need one version of the truth to run their business from inventory to allocation of marketing funds to sales force compensation. To make your sales forecasting most effective, you can’t stop at the standard measures, you also need to focus on end-user demand forecasting, operational forecasting and long-term contracting. Leading companies should look to understand how to incorporate all three to maximize forecasting accuracy.

Kim Frazier: Senior-level executives have an increasing need to be able to forecast end-user demand in response to rapidly shifting customer preferences and competitive threats, but it is not easy. First, there is limited availability of sell-through and point-of-sale data from distributors and retailers both inside and outside of the United States. Second, the data you can get from various third-party subscription services can be inconsistent or even contradictory. Finally, it is hard to close the gap between the data you can gather and procure and the final answer of end-consumer demands. Other industries have faced similar problems and they have come up with a myriad of solutions. These organizations focus on four main areas of improvement, using joint business planning efforts and partners to get more and even more relevant data, using prioritized rules and better forecasting analytics to reduce data conflicts and bridge the gap between the data you have and the final forecast, streamlining the forecasting process to create one coherent picture across the organization and implementing a final forecasting engine to conduct the analyses and publish the results.

Srini Bangalore: From an operational perspective, successful semiconductor companies focus on three major capabilities. First, companies have found it is critical to forecast at an acceptable level of granularity, so that it is operationally actionable information, whether that is at the product family level or at the SKU level. The next step for operational forecast success is cross-functional collaboration. A leading practice, often referred to as integrated business planning, is to have sales, supply chain or operations and finance reach consensus on the range of forecast so that the entire company is executing to one operating plan. Finally, leading companies collaborate with their top customers and partners. This is certainly challenging because customers frequently change forecast and it is difficult to control partner behavior. The result is your forecast can be off significantly and cause large inventory fluctuations. Companies that implement all three of these operational capabilities can see improvements in their forecast accuracy of 20% to 40%, inventory reduction of 10% to 30% and on-time delivery performance improvement of 25% to 50%.

Kim Frazier: Deal-driven sales models have been another area driving volatility for semiconductor companies. The ability to fix forecasting through contracting pushes some of the risk upstream and gives you more stability. We are seeing companies move toward locking in 90 or even 180 days of stability in their volume, as well as other pricing aspects so they are contractually locked into more certainty. This also helps to minimize cost fluctuation as volume becomes more certain.

Scott Angel: While there is no magic bullet, improved forecast accuracy is within reach. By applying focus to end-user demand forecasting, operational forecasting and long-term contracting strategies, you can greatly enhance your forecasting accuracy and in turn profitability.

Related links

  • At the core
    Examining issues and trends keeping semiconductor executives up at night.
  • Technology
    Learn more.
  • Rising tide
    Exploring pathways to growth in the mobile semiconductor industry.

Share this page

Email this Send to LinkedIn Send to Facebook Tweet this More sharing options

Stay connected