Know-How Exchange : Research/Metrics

Hi Omar,

Part of your answer can be found in responses to your previous questions, and in this one:
https://www.marketingprofs.com/ea/qst_question.asp?qstID=29298

Basically, people who choose the neutral option mean actually that: they have no strong opinion about your company, and should therefore not be counted as 'satisfied'. maybe 'content'? To really interpret this data, you would have to look at how this group reacts to other questions, such as intention to repurchase.

The key question is what you are going to use this for...I would not count them as 'satisfied' for eg performance bonuses. Whether or not these neutral consumers would switch away from the company depends on the degree of competition.

The best way is to total all responses except those that are neutral and figure the percentages that way. Example:
10 say great (or, whatever your highest cateogory is)
10 say good, etc.
26 say neutral
5 say bad or fair
5 say poor, etc.

Total 10, 10, 5 and 5 for a total of 30 responses (don't even count the 26 neutrals).

So 10/30 say great (33%); 10/30 say good (33%); 5/30 say bad (16%); 5/30 say poor (16%).

That gives you information about the people who actually have a strong opinion or reaction or experience with your product or service.

I can see how you might want to omit the neutrals in some instances, as demonstrated above, but in other situations it can be very helpful to keep them in your calculations. It can help provide you with an understanding of those consumers who may still be influenced, so, for example, you may want to create cross-tabs on these "neutral" individuals to observe ways in which they might be influenced (depending upon the other questions you ask). It can also be helpful to include these respondents in tracking studies, when you are trying to determine if the neutrals are moving into positive or negative territory, or if they are expanding as a slice of the pie.

Weight the responses: On a 5-point scale, you might give the top box a 5, second a 3, middle (neutral) a 1, then -3 and -5 to the bottom box (only do these last two if you want an overall impression).

Change the weighting according to your needs and the realities of the category (i.e., you might not count the neutral responses at all if the category is well-established and there are meaningful substitutes/competitive brands.

Best of luck!
-Steve

I don't understand what the problem is. If people are neutral, isn't that a valid opinion? What are you trying to prove that would have you considering what to do with those responses? Top-2-box score is not affected by what you do with the neutrals. Maybe there's an issue I don't understand. Can you explain?