Digital camera market has enjoyed tremendous growth since it was introduced in the photographic industry. Starting in 1998, price has been falling rapidly. In addition, the development of CMOS allowed digital camera to not only capture professional market, but also enter consumer market. It was estimated that in 2006 forecast would peak with 63% penetration rate for digital cameras in the US. After 2006, the growth rate was expected to fall negative.
Product lives for digital cameras had been shortened. While an average life was between 17 and 22 months, high-end, feature packed products had the shortest life cycle.
However, the manufacturing lead-time for digital cameras was long. Taking Leitax for example, its manufacturing lead-time was beyond 13 weeks. Therefore it was very important to have an accurate forecast to avoid costs of write-off and obsolescence and loss of sales.
The growth saturation has brought supply chain challenges to Leitex. The maturing of digital market also meant more product launches and more overlap features of products to be managed.
It became harder to achieve accurate forecast. In addition, having expeditious demand and production planning, flexible upside and downside supply, and achieving few holding inventories were more vital as errors can impact financial performance severely and lose its customers.
By the end of 2002, Leitax had bared huge loss through inaccurate forecasts and poor planning of three models. One model was delayed, one model was out of stock, and the third one was having sluggish sales. It was estimated that the cost of delay, the lost sales and excess and obsolescence costs accumulated to .
However the maturing market also brought Leitax an opportunity. With technology development, cost of digital market production was decreasing. While there were more products in the market, many of them could be close substitute of each other. While every company in the industry faced the same supply chain challenges, If Leitax successfully could manage its supply chain efficiency, it could expand its market share by capturing sales of competitors’ products.
Traditional forecasting was ill defined. The sales group made forecast on new product introduction and midlife product replenishment, the operations group used the forecast to drive its supply chain, the finance group evaluated the forecast to see if they can meet financial plans and urged adjustment if needed. However these functional groups could manipulate forecasts based on their own incentives and interests. The three key actors in the forecasting process had conflicting goals: the sales group’s incentive was on commission on sell-in to the resellers; the operation group wanted the most stable and desirable forecast because it was easier to manage; the finance group wanted to meet expectations on financial targets such as revenue and profitability. Not only these forecasts had large discrepancies, the biases and ignorance of real market information each group had significantly reduced forecasting accuracy.
As an outcome of inaccurate forecast, supplies mismatched market demand, resulted in loss of sales, and inefficient and disorganized supply chain operations. Due to ill-defined and inaccurate forecast, Leitax experienced huge financial loss in 2002. As introduced previously, he delay of a model resulted in lost sales, the opportunity cost of warehousing management and arranged capitals that was waiting for the launch that summed up to $19.5 million. The stock out on the second model reduced gross margin for about $4.5 million. The obsolescence costs on the last model was about $2.5 million. No matter how sounding the financial strategy was, if its forecast was not accurate, an enterprise like Leitax would have difficulty in achieving it financial targets, or would even fail to survive.
Therefore, it is an urgent demand to have a consensus forecast system that could incorporate each group’s needs and priorities: to enjoy the sales growth that would achieve financial target such as high profitability, and having an efficient and easy-to-manage production and operation system.
Before, the concept of demand forecast was to serve the key functional groups in achieving their own interest. Facing the new challenges, forecast needed to be more accurate. And therefore it needed a new concept that is to have a consensus forecasting that would accurately reveal market demand and align the needs of key actors in the forecasting process. Leitax implemented two specific changes in forecasting process. The first one is to switch the focus from sell-in to sell-through and second one is to ignore capacity constraints.
The mission of the Forecast Re-Design Team was to reduce expenses and risks associated with production and costs of procurement, inventory, quality control and technical infrastructure. According to Fowler, there were four objectives: “(1), to reduce inventory levels across the supply chain; (2) to increase velocity and accuracy of planning information throughout the supply chain; (3) to increase supply upside and downside flexibility; and (4) to improve on-time performance to customers.
The projects require individuals who have good technical skills and forecasting and industry knowledge. But more importantly, it needs people who could help implement the new changes. To be more specific, the qualities are the leadership abilities to shape an organization and know-how to effect senior management to modify the processes.
Leitax applied several steps in change their forecast process. In the early consensus forecasting meeting, the Redesign Project team realized that variances between the forecasts was largely due to the range of business assumptions about price and promotions. Therefore the starting point of the consensus meeting was developing a business assumptions package (BAP) that included information on price, market trends; competitor’s models and marketing plans. The BAP is critical to overall process because the functional groups (PPS, Sales, and DMS) used it as the basis for additional forecasts.
The product planning and strategy (PPS), marketing and DMS group put the information together in BAP and updated it monthly.
The second step was to create three forecasts, PPS created a Top-down forecast based on estimation of industry demand, historical and current trends of Leitax’s market performance and the current portfolio. The Sales would create a bottom-up forecast based on information of what is happening in the distribution channel. The DMS group prepared the last forecast of sell-through based on statistical interpretations of historical data of sales.
As the third step, differences between three forecasts were evaluated in the Consensus Forecasting meeting to achieve consensus forecast. Though finance group did not participated directly in forecast process, their opinions were addressed in the consensus forecast during meeting. In addition, other topics such as new product introduction, discontinuing of current products, and feedbacks were also included in the meeting. Once the final consensus forecast was achieved, the finance groups evaluated it and approved it after revising it carefully with Marketing and Sales groups. The Final Consensus Forecast was sent to generate the MPS. DMS also used a forecast accuracy metric to evaluate the performance of this forecast.
These steps changed “stovepipe forecasts” into a system that gave consensus. By doing so, forecast accuracy was greatly improved. As a result, inventory turns increased from 3.5 in previous year to 5.4 in 2004 Q4; average inventory decreased from $142 million to $101 million. Excess and obsolescence costs were reduced from $3.1 million (average in 2001-2003) to zero in 2004.
A Single Forecast-Driven Planning Model allowed Leitax to collect assumptions about future product demand and performance from different areas of the organization and to summarize these assumptions and their implication. It also allowed key players to align their needs and priorities. Therefore it helps to achieve a more accurate forecast. Forecast accuracy is very important to manage inventories, to capture sales when there is demand and also to avoid obsolesce and write-off costs. All those are directly related to financial and revenue implications. Thus a Single Forecast Driven Planning Model is highly important to financial performance of the enterprise.
The differences between actors in forecasting process was achieved through a mechanism created by Fowler to allow each group to formally communicate on forecasts and evaluate it through steps. As discussed in the last question, consensus forecasting meeting allowed all parties to understand the assumption driving the difference of forecasts and summarize them in a single document. In addition, there were multiple opportunities inherited in the mechanism for evaluating and revising the forecast to make sure needs from all parties were addressed collectively.