Wednesday, November 18, 2009

IKEA - Proactively Fixing Stock Outs - Is it applicable to others?

Steve Banker writes about the in-store logistics at IKEA.  He describes what they have done to proactively fixing out-of-stocks by observing the sales rate.

Because what is sold is known (POS data) and what comes into the store is also known (WMS data), very little cycle counting is done. However, IKEA’s system is able to catch anomalies. If the system expects a certain volume of a particular product to have sold during a two day period, and much less has sold, the system will direct in-store logistics to go to the location and conduct a manual stock check.

I suspect that it is slighly more sophisticated and takes into account the day of the week. This is a low hanging fruit that most retailers can implement for their highest volume basic items - items that don't have significant seasonality and are not on promotions.

*But* in many consumer goods, these type of items are less than 5% of the SKU count and a slightly higher percentage of the sales. For most other SKUs, this method has been tried and has been found lacking. Some retailers look for "zero scan over a few days" to identify products that need to be checked out.  Other retailers look for week over week sales decrease to identify problem products. These methods result in accuracies of less than 10%, leading to a lot of wasted time and effort and also erodes the trust in the system.

A more sophistcated approach is needed.  An automated day-SKU-store level forecast and the sales (POS) on consequtive days need to be compared to identify the problem products. This has proven out to be 6 to 8 times more accurate.



  1. I will not concur on it. I think warm-hearted post. Especially the title attracted me to review the intact story.

  2. Nice fill someone in on and this post helped me alot in my college assignement. Thank you as your information.