“Forecasting is only for the Fortune 500 and other large enterprises.”
In our years of experience, we’ve seen everybody from a one-warehouse online retailer to an eight-distribution center third-party logistics provider benefit from forecasting.

“We have a legacy/home grown enterprise resource planning system (ERP). It’ll cost us a fortune to upgrade to the latest and greatest ERP to utilize an Inventory Management Solution.”
You are correct: It may cost a fortune to upgrade to the ERP system du jour. However, RockySoft is compatible with virtually all ERP systems and back end databases. Therefore, it is possible to extract the necessary data from your older system and produce excellent results without having to upgrade. RockySoft has extended the life (and investment) of many legacy ERP systems in the past.

“It takes a trained mathematician or statistician to operate sophisticated forecasting and replenishment software.”
It certainly wouldn’t hurt having a mathematician or statistician operating the software. However, RockySoft was designed by a mathematician to be used by anyone who is comfortable using the computer and can learn a few fundamentals about forecasting and replenishment planning. Basic training at RockySoft is measured in hours-not days, weeks, or months.

“I’ve heard the old maxim ‘forecasts are always wrong’, so why bother forecasting in the first place?”
We’ve heard that maxim before too. In fact, there’s a lot of truth in it. However, by creating a forecast and measuring its accuracy over time, you can begin to plan the appropriate buffer or safety stock. Utilizing a little science and statistics, you can even begin to predict with some accuracy the amount of stock it will take to achieve a particular customer service level goal or fill rate.

“The more sophisticated a forecast algorithm – the better.” Leonardo Da Vinci once said, “Simplicity is the ultimate sophistication.”
Over the years at RockySoft, we’ve found that while there is some benefit to having a sophisticated algorithm around, they’re not necessarily better. Often, simple models can provide you with as good or better forecasts than the complex ones. Our strategy is to provide you with a number of simple and several complex algorithms, provide you with their predicted error rates for the historical data and let you choose the one you feel best fits the needs of your business.

“The more history I have on an item, the better the forecast will be.”
In today’s rapidly changing environment, demand patterns are constantly evolving. As a result, having more data doesn’t necessarily mean better forecasts. We believe that the more recent the data, the more likely it will be in producing valuable forecasts. In most industries, three years or thirty-six months of history is plenty. In some high tech industries, even less history can be used effectively due to the short product life cycle.

“There is a magic model or forecast algorithm. It is just a matter of finding it. Once you find it, it can be used for every SKU and even every category.”
In historical time series forecasting, every data series is unique. Consequently, there is not one algorithm available today that will provide a good forecast for each and every SKU. At RockySoft, we provide more than a dozen algorithms to analyze all of the SKU/location combinations automatically each day. This provides a forecast and historical error rate for each SKU. That way, you can determine which model produces an acceptable error rate for your business.

“A good forecasting solution will solve all of my forecasting woes!”
This is akin to saying that if I buy the best golf clubs, I will be able to beat Tiger Woods. Unfortunately, it does not work that way. As with golf, there is a bit more to forecasting. Aside from having a good solution or ‘tool’, one must be able to analyze data for problems, understand the fundamentals of forecasting, put solid processes in place, and determine whether a model or replenishment suggestions appear to be correct. If not, there is the propensity to end up with ‘garbage in and garbage out.’

“Historical time series models do not provide the most accurate forecasts.”
Time series models assume that historical demand patterns will continue into the future and can be very accurate. RockySoft utilizes time series modeling to provide an automatic baseline forecast for each and every SKU. While this provides a good forecast for many of our clients’ SKUs, it is just the beginning for others. RockySoft supports collaborative forecasting for those companies who have sales reps with an intimate knowledge of the customer. Additionally, RockySoft utilizes an ‘operational’ forecast line for replenishment that can account for special events (promotions, etc.) as well as insider knowledge not available to the software.

“We have more than 50,000 stock keeping units (SKUs) to keep track of across four warehouse locations. I can’t create forecasts for all of them?”
It would be difficult for a human to perpetually create forecasts for that many SKUs. RockySoft utilizes a RAM-resident engine to create a forecast for each and every SKU/location combination every day. Additionally, RockySoft utilizes an exception based management system to alert you to any SKU that falls outside of a variety of user-defined parameters. That way, you can spend your valuable time reviewing only the SKUs that need the most attention.

“Safety stock is only calculated to cover demand variability. What about supplier variability?”
The need for safety stock (buffer stock) to compensate for uncertainty arises from two primary areas: demand variability and supplier (supply) variability. At RockySoft, safety stock is dynamically calculated using lead time, forecast model error, forecasted usage, and service level. That addresses demand variability. To address supplier variability, RockySoft uses a sophisticated lead-time management module that can insert an actual or padded lead time instead of the static planning or average lead time that ERP systems generate. We can help you reduce lead-time creep and hold your suppliers accountable.