Promotional SKU Vs. Like modeling

[Former Member]

Dear DP Experts,

I see a strong business case for product life cycle planning and promotional planning in my project. While I appreciate the theory behind various cases of promotions and life cycle planning, I wish to be advised of any post implementation headaches (e.g. too much data maintenance, reporting complications,. additional time spent))  that might arise with following options.

- I treat promotional SKU as a new SKU and I simply copy (RLGCOPY) some reference data from from its "regular" counter part. I can use some copy factor based on expected additional demand and get going.

- I create a like profile when promotional SKU is launched for a short period.

- I create a phase-in / phase-out profile when regular SKU is discontinued to promote new SKU.

Assume a typical CPG firm, largely make to stock. 

More generally I wish to know, in a business where promotional SKU's are almost a regular characteristic of the business, what creative ways exist in DP to model the same with little data maintenance efforts (trade off !)

Thanks

Guru

[Former Member]

Guru,

Did you devise a strategy for managing promotional SKUs?  I am facing a similar situation with my current project.

[Former Member]

Hi Bob,

Heres how I am going ahead (strategy)

What am I calling a Promotional SKU ?:

1) A new SKU of lesser price, new packagaing, more quantity or some form of incentives for consumers like coupons and lotteries and another item being bundled with it, WHEN there already is a regular SKU that is not discontinued in same or other markets. This may or may not be supported by a promotional event (media).

-In this case I use a like profile to estimate my likely demand and plan the first 3 months of sales as 100% promotional demand and thereafter being 100% regular demand. So both SKU's exist. cannibalization group can be set up by the planners but this is not "forced" on them

2) Ditto1) when a regular SKU is discontinued from an effective date in future.

-In this case I use a phase in and phase out profile to estimate the likely sales of new SKU and plan the first n months of sales as 50% promotional demand and 50% regular demand and there after as 100% regular demand.

3) A bundle of regular SKU's in stock that are sold at a discounted price either to trade or to consumer. In this case the product is sold as a new promotional SKU. There is no way to know the constituent SKU's in the sales order(no sale order BOM's, no DP BOM's being used in my case). These are ad-hoc promotional sales just to get the stock out from supply chain. This can be company's plants, warehouses or regional DC's or even stock at retailers.

-In this case I plan the entire demand as promotional demand for this promo SKU. The original demand planned for constituent SKU's is not adjusted because forecast error will have to explain the fact that the bundled SKU demand was not planned to begin with. The condition here being planners are made aware of such promotions in advance. There is a bit of a catch here. It is likely that planners are unaware that such SKU's is being produced out of "thin air" almost in the last minute.

In all these cases I intend to separate promotional sale from total sales (history) in my planning table. However there is no guarantee here that a promotional SKU is treated as promotional sale in sales order. i.e. there may be no distinguishable parameter in sales item table that SKU A is regular sales and SKU B is a promotional sale

I am also not sure how practicable is this in the long run. You may have to devise your own strategy. Hope these guidelines help.

It may help to have a separate key figure in planning table to record promotional sales distinct from actual sales history. How you update it is something you need to work out. It is possible that what is planned as promotions is not necessarily identified as promotional sale when the actual sale happens. So for history adjustment manually, this may get a bit tricky. In standard DP, the planned promotion quantity (amongst other things) is what is reduced from actual sales (with some cosmetic adjustements) to get a corrected history but this may not be reliable enough !

Thanks

Guru.

[Former Member]

Thanks for sharing Guru.  Very helpful.