sample="quota" bates="516801931" isource="rjr" decade="1990" class="ni" date="19961112" Reggie, Harvey E-mailed this over for our meeting Thursday. Check your E-mail for a clean copy. ABP Subject: MM Pack Test Action Standard November 12, 1996 From: Harvey Gordin In thinking more about an appropriate action standard for the upcoming MM Packaging study, I have come up with three possible approaches. I would like for all of us to settle on an Action Standard that does a proper job of controlling risk while at the same time does not limit us in exploring a potential opportunity. A meeting to discuss this will be showing up on your calendar shortly. Some of the proposed Action Standards are described below. We can discuss these and your alternative suggestions at the meeting. In each of these approaches the focus is on Purchase Intent as it is assumed that we will not accept any downside risk in "Taste", i.e. %70+. Given the above assumption (Which we can certainly discuss if you do not agree), the following options are available: 1) Most Conservative Approach - No Downside Franchisee Risk To be considered for implementation, ST-Pack must have Parity or Better %70+, Parity or Better % Top 2 PI, usually tracks w/%70+-do we need it? and Parity or Better % Bottom PI among Franchise Smokers, AND significantly greater % Top 2 PI, and or Parity or Better %70 + -(among Competitive Smokers, I think we should use the either/or standard) - we also need to include something on perceptions among Competitive Smokers. If the above Action Standard were utilized, the ST Pack would not be implemented given any of the following set of results: Obviously there are additional potential outcomes. However, these examples illustrate that under this most conservative approach, Franchise Risk will not be tolerated regardless of the potential competitive gain. 2) Less Conservative - Place a cap on Franchise Risk A less conservative approach would be to set a "cap" for Franchise risk based on purchase intent. For example we could agree that a decline of % Top 2 Box of more than 5 percentage points (or 2 or 10) will not be tolerated regardless of the Competitive Upside. An alternative to the above would be to simply accepting differences in % Top 2 Box at 90% confidence. This would result in making it harder to identify a difference, while at the same time, allowing smaller differences to "slip through". 3) Even Less Conservative - Create a "Risk/Reward" Ratio The final approach would be to set a limit on the declines in Franchise Purchase Intent based on the increase in Competitive Purchase Intent. For example, we could stipulate that regardless of how large the drop in Franchise PI is, that drop will be tolerate if the gain in Competitive PI is 2 times (or 5 times or 10 times) greater than Franchise loss. Please give this topic some thought. I hope to come to an agreement regarding the action standard at our meeting. This test is currently scheduled to mail out on January 2, 1997.