The replenishment forecast series settings screen is where you can configure the settings for the replenishment forecast demand that is shown in the Order Schedule.
Access this screen by going to Admin --> System Configuration --> Replenishment Forecast Series
The main use for the replenishment forecast is the conversion from the demand forecast you have in Forecast Manager into the day-by-day demand forecast laid out in the order schedule. In most installations of StockIQ, there will be only one Replenishment Forecast Series, usually named "Independent Forecast"
To configure your replenishment forecast series, click on the entry, and then adjust the settings in the available boxes:
Source Demand Forecast Series - This tells StockIQ where to source the data for your replenishment forecast series that will appear in the Order Schedule. Typically, this will be your Independent Forecast Demand Forecast Series from Forecast Manager, usually called "Independent Forecast"
Consumption Demand Series - When performing Forecast Consumption, which demand series should be compared against for looking at firm demands that have consumed our forecast? Again, in most situations, there will be only one demand series, usually labeled "Independent Demand" or occasionally "Actuals"
Conversion Bucket Method - The Demand Forecast Series coming from the forecast side of StockIQ is always calculated and saved in weekly or monthly buckets. Typically, this is then converted to a daily replenishment forecast according to the monthly or weekly Demand Pattern. The "Auto" method will convert a monthly forecast into a month's worth of daily buckets, a Weekly forecast into a week's worth of daily buckets etc.
In certain rare cases, you may want the monthly forecast placed on a single day, e.g. the entire monthly forecast represented as occurring at the beginning of the month, or the week, but this is very uncommon.
Conversion Rounding - When converting from the monthly demand forecast to the replenishment forecast, when you have very small numbers, e.g. a demand forecast of < 1 per month, this can result in extremely small daily quantities, which in turn can lead to unnecessary replenishment suggestions, such as when StockIQ predicts you might go *fractionally* below zero, and therefore need to purchase more.
To prevent this, when converting from the demand forecast to the replenishment forecast, StockIQ can either:
- None - Do not modify the demand forecast, allow fractions to flow through.
- Drop - Drop all fractional quantity (e.g. 0.9 becomes 0, 1.8 becomes 1, etc),
- Round - Apply traditional rounding
- Front Load - Fractional portions are accumulated and then grouped into the earliest available bucket, so monthly forecasts of 0.4, 0.4, 0.2 would turn into 1, 0, 0, and then split into the replenishment buckets.
- Back Load - Fractional portions are accumulated and then grouped into the earliest available bucket, so monthly forecasts of 0.4, 0.4, 0.2 would turn into 0, 0, 1, and then split into the replenishment buckets.
NOTE: An alternate approach to conversion rounding that can be effective, especially for low volume items where this is important (e.g. forecasts of <= 1 per month), is to allow order policies to take over instead, in that Min/Max or Sporadic Sales policies are often more appropriate on low volume or sporadic items anyway, and forecast is largely ignored by these policies.
Consumption Level - Defines the level at which forecast consumption is done. For most organizations, you will want to set this to Item-Site level consumption.
If your hierarchy includes customer-facing attributes, such as specific Customer-Ship-To, or a category such as the Customer Category 1 to make forecast groups, then you may wish to also make your replenishment forecast consumption happen at a level consistent with your customer groupings, such as a forecast group. In this case, if you have a large customer, such as Walmart, who can possibly consume your entire forecast with a large order, the effect of that large order on consumption is shielded from your other forecast groups. This can be useful, since even if Walmart places a large order that consumes all of the Walmart expected forecast (or more), your remaining customers are independent of this, and can still be expected to consume their forecasted quantity, and StockIQ can make sure you are still planning for this with the customer or category level consumption.
NOTE: This setting has no effect when your consumption option is "Auto-Linear"
Consumption Order - When doing consumption, either monthly, weekly, or daily, this defines how StockIQ runs the consumption logic. For example, if you are set to monthly consumption and receive an order for 100 units on the 15th of the month, should it begin consuming from the first of the month forward, or should it consume from the 15th and walk outward in increasing number of days?
- Middle-Out - this is generally the most sensible approach. Middle-out means that we consume from the point of the sale outwards to do the consumption, much like ripples on a pond. If a sale lands on the 17th, then we look for forecast to consume on the 17th, then 18th, then 16th, then 19th, and so on, much like ripples radiate from the center of a rock thrown into calm water. Consumption stops when we reach the limits on consumption, such as -10 / +10 days as in daily consumption, or at the week or monthly borders as in weekly or monthly consumption.
- Early To Late - The early-to-late method starts by finding the very earliest period that it is allowed to consume given a demand that must be consumed. Similar to the middle-out example above, if a sale lands on the 17th, and consumption is allowed up to 10 days before, then consumption will start from (17-10 = 7th), and then proceed forward until the demand has been completely applied to the forecast.
Consumption Option - This defines the rules on how your forecast is consumed.
- None - This option only applies if manual forecasts are being auto-uploaded into the system and consumed via some external means. It is generally not applicable except for special cases.
- Auto Linear - This method of consumption assumes that the forecast is consumed in a linear fashion as the period goes on, and is insensitive to the actual patterns of orders received. The advantage of this is greater stability due to the consumption proceeding in a day-by-day manner. The downside is that there is no response to actual ordering behavior on the part of your customers. however, any error is in the conservative direction, e.g. you are more likely to slightly over-order than under-order, making this a safe, stable approach that works well for many organizations.
- Monthly - Consumption is done within monthly boundaries. Demand orders received within a month are consumed only within that month, not beyond. This has the advantage of being responsive to actual orders received, but has the disadvantage of adding more jitter to your plan. For example, if you reach the end of the month and NONE of your forecasted orders have occurred, StockIQ will still be condensing your entire monthly forecast into those last few days, expecting that forecast to come true and be consumed by the predicted orders. This can lead to a large order suggestion. However, when the month rolls over, that un-consumed forecast can then disappear, and along with it the order suggestion, which can be an unstable state of affairs.
- Weekly - Same as monthly, just in weekly boundaries.
- Daily - This allows for consumption of your forecast, but within a specified number of days on either side of the demand order. This allows for more flexibility and less instability due to month boundaries, e.g. if an order falls on 10/31 vs 11/1, there is not a large difference in how the consumption is done. For this reason, the daily consumption method is a good choice for active consumption if the Auto Linear method is not appropriate for your business, and we generally recommend using a relatively wide range of days for the consumption, such as +/- 7 days or so.
Daily Consumption Days Back / Days Forward
These allow you to specify the limits for daily consumption, e.g. what are the ranges back in time and forward in time from the date of a demand that StockIQ will run the consumption. This option appears only when "Daily" consumption option is specified.
Max Consumption Months/Weeks Forward
When doing monthly or weekly consumption, you have the option of allowing the consumption to cross the monthly or weekly border going forward in time if desired. This can help prevent the "stack up" of leftover forecast at the end of a month or week. No "months back" or "weeks" back option is available.
Include in Daily Usage - This option specifies if the replenishment forecast series should be included in daily usage calculations. For your primary replenishment series, this should always be enabled.
Allow Negative Forecasts - Allows the notion of a negative forecast (e.g. planned credits or returns), or if the forecast should always be capped at no less than zero.