A supplier with seasonal products "found" $7 million in incremental sales by using Shiloh.  Here's how... A line of seasonal products were in distribution in about 800 stores.  Sales were down significantly in the sub category and the buyer was seriously considering cutting the line from all Walmart stores to make room for more successful items. When the 800 stores were aggregated, the entire category did appear to be down. Using Shiloh, the category manager was able to look at individual store sales and determined that there were in fact about 100 stores that were selling quite well.  Once the “good” stores were identified, he was then able to use Store Fact Clustering in Shiloh to find out what if anything the good stores had in common.  Store fact clustering revealed that the “good” stores had several commonalities.  Many of the stores were Coastal, near a military installation, and were in states that had laws specific to these products.  Using this cluster of Store of the Community traits as a “profile” for successful stores the category was then able to use trait matching in Shiloh to identify stores not in distribution but matched the trait profile.  A Store list of recommended stores was then generated and presented to the buyer. Based on the analysis, the buyer decided to drop the 700 or so “underperforming” stores and in the end added an additional 1200 stores that fit the trait profile.  This equated to an incremental $7-million in sales for the customer.

Sun Products implemented Shiloh and within the first 30 days identified $32,000,000 in lost sales due to store distribution, replenishment setting, and out of stock issues, which were also causing forecast accuracy issues. Standard Shiloh "zero sales reports" pointed out item/store level issues that needed correcting and automatically sent these to store merchandiser reps for resolution. Shiloh comes with a report where a product is Traited and Valid but there are zero sales and zero OH’s showing in the stores and or DC’s.  This report is automated on a daily basis shows items and stores with zero OH’s and zero sales.  Habitual problems are identified (number of occurrences over a time period, i.e. 7 occurrences in the last 10 days).  

Church & Dwight now has two additional days to concentrate on strategy instead of report compilation.  Church & Dwight had to produce six weekly Walmart Supplier Stats Reports. These Stat Reports took one person 3 days to create beginning on Wednesdays to complete by 9:00 am on Friday. By automating these reports this analyst now has an additional two days each week to focus on forecasting and instock.