Industry analysts, big-time consultants, and your peers are all talking about technology, digital transformation, and the future of the supply chain. It can seem like a lot of noise given the day-to-day pressure you feel while working to ensure that inventory is on hand and positioned where it is supposed to be. Every. Single. Day.  Maybe this is the crux of the issue. With all you have on your plate, are you aware of the signs that it is time for a change?  Here are a few things to keep your eye on.

#1 Manually intensive end-to-end forecasting and planning process.

Your team spends countless hours finessing your forecast with manual overrides and you still cannot exceed 60% forecast accuracy. And the number is even worse for items with intermittent demand.

You may be ready for a fresh look at how you are forecasting. Your supply chain software should help you forecast over a wide array of items (fast movers, intermittent, slow movers), across the entire network (even 2 or 3 tier), and provide target inventory levels (safety stock, reorder point, cycle stock). Making this change could help you achieve the service levels you desire and provide a continuous feedback loop for incremental improvement.

#2 Forecasts are only in monthly buckets, so daily/weekly trends are lost. 

Your system is not capable of analyzing data or consuming forecasts at the weekly or daily level, nor at the “Ship To” location. With the right technology, you may discover micro-trends that occur within the month and be better able to capture that demand and produce demonstrably improved forecast accuracy. This is especially powerful if done automatically at the “Ship To” location level.

You might be ready for looking at ways to sense demand in the near term at a granular level which can be used to enhance the overall outcome towards the customer.

#3 “Best Fit” forecasting turns out to be “whatever fits”.

Forecasting based on best fit from a limited selection of formulas is prone to overfitting. It does not really model demand. It simply jumps from model to model, using whatever worked last time, without correctly understanding demand drivers.

Consider statistical “order-line” based modeling to help understand the drivers of volatility drivers, both natural (trends, seasonality, causal factors, and lumpiness) and organizational (demand shaping promotions, new products, forecast bias, and the bullwhip effect) to provide you with a reliable forecast.

#4 You are unable to make an effective inventory to service trade-offs.

You don’t really know what investment is required to hit the target service level your executive team desires. You are stuck with a “one size fits all” inventory policy across all SKUs and locations, stocking product which sells once a year when you could use the capital (or space) for products that sell thousands of units per month.

You ought to be able to perform “what-if” scenarios to evaluate multiple inventory optimization scenarios and choose your inventory sweet spot. Imagine what the value of getting a detailed breakdown of inventory targets, safety stock, and recommended service levels by individual item-warehouse combinations would be for your team and quality of work-life.

#5 Cannot handle multi-sourcing and simultaneous sourcing.

Products outsourced from multiple locations force you to manually split aggregated forecasts and layout separate inventory plans to meet the target service level. Not bad for a few items, but a serious problem for thousands of Item Location combinations which in turn can have millions of components that go into them.

Multi-sourcing allows you to deal with products that must be sourced from multiple suppliers. You fix the percentages sourced from each supplier and the system calculates the total inventory needed at an aggregated level to achieve your desired service level and roll out replenishment orders for each supplier. Plus, it understands that your suppliers have different lead times.

Learn More: Machine Learning: Optimization and Community Intelligence – A Wiser Forecasting


Even one of these five signs should tell you to take a long hard look at your technology. But the big question remains – is it worth it?  Updating or replacing planning software can be time-consuming and requires a strong cross-functional team. You should probably know how deep the water is before diving in.

Swifcast Powered by Arkieva is a simple business case/ROI calculator that can provide the insight you need to determine the value of updating your approach to supply chain planning. With Swifcast, you upload historical supply chain data and receive a forward-looking forecast. This forecast is based on machine learning algorithms and not only tells you what to order and where to place it, but it also spits out a projected ROI if you follow the recommendations. With this ROI, the corner office will feel comfortable with the investment, and you can retake control of your planning process. You can register for Swifcast here or complete this form to speak with one of our ROI consultants.