Archive for January, 2007

NRF Show 2007

January 16, 2007

The Accelerated Analytics team had the opportunity to attend “The Big Show” today in New York City.  If  you are not familiar with the show this is the annual trade exposition sponsored by the National Retail Federation.  Overall the show is a good source of industry news and provides an opportunity to spend time learning about new technologies and vendors in the market.  The show floor can be a bit overwhelming with thousands of vendors.  This year we were struck by the number of empty booth spaces as compared to prior years.  It would seem the declining attendee rate has impacted the NRF like most organizations although this is still a very big show.   It is very hard to gauge who is attending but we had the opportunity to talk with representatives from Gap, Jo-Ann Stores, The Andersons, Wal-Mart, and saw many other retailers so it seems the usual suspects are in attendance.  It does seem most of the attendees are part of the IT, store operations, or supply chain teams.  One conversation we had with a fairly good sized retailer regarding their supply chain vendor collaboration was a bit surprising.  This vendor told us they make weekly product activity data available to their suppliers but only a handful really use the information.  This was a surprise because as we work with vendors most tell us the opposite is true – they request the data but their retail partners are not filling that request.  In fact something in between is probably true.  Our experience tells us retailers are generally making the data available but in many cases not in the format the vendor would like to receive so they are challenged to do much with it.  This is unfortunate since there is much that can be done with the data.  Hopefully more productive programs will be put in place in 2007.

If you are in NYC and have the opportunity to attend the show make time to visit the Microsoft booth. (full disclosure we are a MSFT partner)  Their booth is huge and filled with solutions for every aspect of a retail operation.  The representatives working the booth are overly technical but if you ask questions about their client work they can provide some very interesting tid-bits of information.  We especially enjoyed our conversation with the Project Real team.  If you want to learn about how a huge data warehouse is deployed at a real customer you need to check that out.

 We dropped our cards into most of the fishbowls we found so maybe we will win something:)

Using Sell-Thru for decision making

January 8, 2007

Sell-Thru is a key performance indicator for vendors and retailers alike.  Sell-Thru allows one to understand the velocity with which inventory is being consumed as it relates to sales.  Because sell-thru is a leading indicator it is also very useful for predictive analysis.

When calculating sell-thru careful consideration should be given to how the formula is constructed.  As with any business metric there is more than one possible answer, but only one is correct in terms of meeting the business users needs.

So ask yourself: what is the difference between calculating sell-thru using all available weeks of sales and inventory data as opposed to using the most recent 4 weeks of data?  Both methods are valid, although they may produce very different results.  The first method will tend to flatten out fluctuations due to promotions.  This is useful if the item is on replenishment and operates within established min/max guidelines.  The second method of calculating sell-thru using the most recent 4 weeks of data tends to provide a rolling snapshot of performance.  This is very useful if an item is  highly promoted and one wants to understand the impact of lift within a given promotional window.

Both methods are correct, but one will be more useful than the other to business decision makers at your organization.  Here are three best practices to make sure your team arrives at the most useful method:

1)  Have simple design sessions with business users to write out on a whiteboard all calculations.

2)  Discuss if the calculation supports the intended business decision.

3)  Adjust the formula accordingly.

4)  Identify low, middle, and upper performance conditions for each metric so exception dashboards can be created.

5)  Document your work in a place all team members can access so there is no confusion on how the calculation is performed, or how the performance conditions are aligned.