- Supply Chain Platform
- Elite™ Enterprise
- Elite™ Healthcare
- Omni™ Retail
- About Us
Know your A, B and Cs
Which products are the most important to your business? A classification tool that allows for a selection of factors based on weight will quickly classify items based on their impact to your organization with A items being the most important, B less so and so on.
An experienced planner equipped with the right tools digs deeper using descriptive analytics with a special focus on A items then onto B items followed by C items. This allows for a better interpretation of the item’s contribution to revenue.
If you have not removed a few C items from your stock inventory in a while, then more time must be dedicated to eliminating slow-moving inventory. Inventory that does not turn over impacts cash flow and profits.
Understand Demand Versus Sales Statistics
It is imperative to capture complete, detailed customer demand coupled with the power to mine this valuable data. Demand represents customers’ willingness and ability to buy product. Sales are a measure of your organization’s willingness and ability to provide product. In other words,
demand is what your customers ordered, while sales are what your organization ultimately sold.
When looking at historical data as a predictor of future events, demand statistics typically offer a more accurate representation. In many industries,
the difference between demand and sales signifies lost opportunity. This is not true for all products in all markets. It will be important to explore why that gap exists and adopt demand shaping and/or supply strategies accordingly. Furthermore, this exercise will lead to a greater understanding
of the factors that influence demand.
How do you balance demand and sales?
If you know how much demand a product will have at a specific price point, you will know exactly how many units to stock to maximize sales and minimize cost.
Categorize Items by Demand Pattern
Visualizing data provides valuable insights into customer buying patterns. Start simple with a univariate time series at the stock keeping unit (SKU) level. A linear graphic will reveal a demand pattern. Non seasonal items will fall into one of four demand patterns: smooth, erratic, intermittent or both erratic and intermittent. Identify the items that fall into each pattern.
Next identify and tag seasonal items. These will need to be managed separately. Ideally, you have access to three years of history to identify a decent seasonal factor.
Use the COV and SD Metrics
Coefficient of variation (COV) and standard deviation (SD) are must-have metrics that planners use to recognize product risk and rightsize safety stock. These metrics measure data dispersion.
Take the following 10 observations: 10,16,15, 5,10,12,10, 8, 7 and 9. The illustration shows the calculation of the SD and the COV for these values.
Step 1 – Calculate the sum of the values: 102.
Step 2 – Calculate the mean: 10.20.
Step 3 – Subtract the mean from the value for each observation.
Step 4 – Calculate the squared value for each
Step 5 – Calculate the sum of the squared values: 103.60.
Step 6 – Calculate the mean: 10.36.
Step 7 – Calculate the square root of the mean: 3.22.
Now let’s multiply these 10 values by 10. The COV remains 31.56% while the SD deviation is now 32.19.
SD factors into how much safety stock to maintain. Although other factors will apply in determining safety stock thresholds, having visibility on an item’s SD provides more context at the SKU level while COV allows a planner to group items with a similar deviation expressed as a percentage.
Demand patterns change over time, so it is good practice to recalculate COV and SD monthly, especially when evaluating items with shorter life cycles.
Measure Forecast Accuracy
The most popular method of measuring forecast accuracy is to calculate the mean average percentage error (MAPE). MAPE is a statistic which offers a prediction of forecast accuracy on past data whereby actual demand is compared to what had been forecasted.
Calculate MAPE at the SKU level over several observations. A high percentage is an indicator of uncertainty hence your supply strategy must consider this metric. Make sure to track MAPE over time and verify whether the error percentage is trending higher or lower.
When buying a product that has a long lead time, how much to buy should be based on forecasted requirements. The potential to misalign supply with
demand is greater with long lead times, so be cognizant of the risk of over/ under buying by considering the MAPE metric.
Improve Forecast Accuracy
Establish a formal internal process designed to improve forecast accuracy.
Are you using the best forecast algorithm? Let your system do the work by evaluating the algorithms that would have resulted in the lowest MAPE over a period of observations. There are many forecast algorithms to choose from. Use simple forecast algorithms to start with because even the simplest algorithm can yield the best forecast.
Your statistical forecast is a predictor of the future based on past demand. Take the time to incorporate business intelligence to improve forecast accuracy. Also, it may be necessary to cleanse historical demand of outliers that have the potential to skew the forecast.
If possible, benchmark your forecast accuracy with companies within your industry. Improving forecast accuracy is time well spent. A study by a company specializing in demand planning showed that for every 1% of improvement in forecast accuracy, organizations report a 1-2% drop in inventory levels.
Hindsight to Insight
Examining why things happened will lead to conversations on what could be. Regular consultations with Sales will add tremendous value. Planners need to prepare for new markets, customers, products and contracts. Encourage Sales to gather intelligence from important customers. In other words, do not limit your inventory forecast to quantitative inputs, seek qualitative inputs as well.
Marketing programs are designed to grow the business while promoting customer loyalty. Marketing programs shape future demand; therefore, the inventory forecast must be adjusted accordingly to better align supply with projected demand. And for the same reasons, planners need to cleanse
historical demand of peaks resulting from a marketing program.
Forecast accuracy affects fill rates, revenue and gross margin. In reality, there are items that can’t be forecasted but are deemed critical. Repair parts
often fall into this category. The supply strategy for these items will require a careful review of demand history to determine how much to stock.
Continuously Reexamine Supply Strategies
Most planners and inventory managers are responsible for hundreds of SKUs. Evaluate A items monthly, B items at least every three months and C items at least every six months.
Your supply strategy review should include the following metrics: order fill rate, MAPE, COV, SD, gross margin return on inventory investment (GMROII), inventory turnover and average supplier lead time as compared to promised lead time.
Armed with these metrics, a planner can determine whether to adopt a procure to stock or procure to order strategy. When procuring to stock, these metrics help planners determine when and how to buy as well as establish safety thresholds in line with the risk of stockout and desired customer service levels.
Your supply strategy review should include the following metrics:
Order fill rate
Mean average percentage error (MAPE)
Coefficient of variation (COV)
Standard deviation (SD)
Gross margin return on inventory investment (GMROII)
Average supplier lead time
Be Aware of Distribution Capacity
Consult with warehouse operations to gain an understanding of labor and capacity constraints. Changes to a SKU supply strategy may impact the warehouse. For example, a decision to order more frequently affects receiving activities and increases acquisition costs. Order by the pallet when appropriate. Also, keep in mind that increasing safety stock on large items may require a slot reassignment.
Reap the Rewards
Demand planning is being recognized as the cornerstone to a mature supply chain. To maximize the benefits of demand planning, establish a formalized cross-functional process within your organization.
Benefits include the following: