Off the Shelf provides a platform for Retailers and Consumer Packaged Goods companies to discuss and gain insights on the pressing problems, trends and solutions.

« February 2013 | Main | April 2013 »

March 7, 2013

Soak Period for Store rollout - A process that needs to be standardized

Rolling out any patch or software update to a chain of stores is always a challenge.  The complexity lies predominantly on the number of instances where the same software package is running.  While there are standard set of challenges for any software update to a stand-alone system, it has the potential to become catastrophic if they were not done with enough attention to details. This is especially applicable for roll outs involving large number of stores.  Many retailers face this problem regularly.  A package that is approved for release and performs as expected in pilot, simply causes unbelievable trouble when it is rolled out to the entire chain. Operations team is bombarded with a sudden spike in ticket volume. What went wrong?  A thorough testing was done before the release, pilot was carefully monitored - so what was missed?

Let's look at the process of rolling out a new application or major software update from a software package like Microsoft Office. For any such major package a beta version is first released to a relatively small set of users. This gives them an opportunity to use and test the latest features with an understanding that it is not fool-proof.  Feedback from these beta users helps to address the initial teething issues, improves the usability of the product as well as helps make decisions about the timing of the release to a larger audience based on the stability of the product.

A similar approach is needed for store rollouts as well. Soak it and measure it. 

'Soak period' is just a fancy term.  It actually means how many stores (beta users) this release needs to be soaked in and for how many days it should be soaked.  The unit for soak period is 'Store-Days'. Soaking in lesser number of stores for a longer duration is the ideal situation to capture maximum number of issues with minimal impact to business.  But 'Time to Market' is also a key factor which would put a constraint on soaking duration.  An optimal duration and number of stores need to be carefully planned for every release, and it may not be the same for all the releases.  For every software / patch that is ready for pilot, Soak Period needs to be identified based on complexity of the patch, magnitude of change in it and the nature of these changes. 

Selection of these beta users or pilot stores is another important element to be considered.  A user who does not frequently or extensively use the product is not a good candidate to be a beta user. A store which does not utilize a functionality that is added as part of the release is certainly not a candidate. 

Few factors that need to be looked at for pilot store selection are listed below. 

• Some stores simply do not report the cosmetic/minor issues. They just live with it.  Stores with relatively high incident reporting trend should be considered for pilot.
• Many retail chains have a variety of store formats.  They vary from a large supercenter with several departments to a small format with one or two departments.   The list of pilot stores should cover all types of store formats.
• Local Laws, regulations, taxes, etc. vary for each geographical location. The pilot store list should have representation from all such unique combinations.
• Few stores in the retail chain may still be using some older hardware. Care needs to be taken to include stores with special or older hardware as well.

Diligent monitoring of pilot stores is the key for successful rollout.  During this soak period, the pilot stores should be encouraged to report all incidents or unusual behavior that they notice.  Operations team should also promptly differentiate a regular incident reported by a pilot store from a release related issue. The project team should prioritize all issues identified during the Soak Period, and ensure these are addressed before the start of rollout to the rest of the stores. It is also important to create a roll out plan, based on the results of the pilot with respect to functional as well as performance requirements. This will ensure a smooth rollout, and a satisfied set of users.

There are a few retailers who perform these activities without having a standard process for doing this diligently for every software patch / release. It is important to make it a standard process for every store rollout. A strong methodology is important to measure the success of these roll outs and fine tune the process, thus ensuring smoother rollouts always.


Posted on behalf of Balaji Rajbabu.