Users' Choice Awards 2005

This year, the changes are limited to the way the survey has been done. For the first time ever, the survey for the personal products was done online, while the enterprise products and services survey continued to be face to face. Otherwise, the methodology has remained the same, with two questions being asked for each product category. The first of the two questions is which is the primary brand the respondent owns for that category of products and the second is, if they were to buy the same category of products in the coming six months, which would be their primary brand of choice.

The Users' Choice is awarded to the brand that gets the maximum votes in the 'likely to buy' category. 

The way the online survey was done merits discussion. This survey was addressed to a database of 6500 developers, system administrators and gamers from across the country. The qualifying criterion was that they had to have access to a PC or notebook while at home. Each person in the database got an e-mail invitation, including a unique URL that expired once the person completed the survey. Once the person accessed the survey online, he could leave at any time and return back to where he left off by clicking on the URL in the e-mail again. Similarly, once completed, the survey could not be accessed again.

Typically, a face-to-face survey of this sort takes many weeks to complete. But by the time we had completed mailing out the invite to the 6500 potential participants (It took about 2 hours to get all the e-mail generated and sent out), about 150 people had already accessed the survey online! And many of them had even completed it!

By a quirk of fate, our survey was launched just before Mumbai was devastated by rains. The face-to-face enterprise survey had to be postponed in Mumbai by about a week. But the online survey got as many as 187 responses from Mumbai during this period!

In all, the online survey was open for ten days and we got a total of 910 responses, which gives a response rate of a shade under 10%. We could have improved the response rate by sending reminders, which we opted not to do.

Needless to say, the cost of running the online survey was only a fraction of the cost of the off-line one, and took less than a third of the time. The key factor to be considered is the response rate. Online response rates are normally in the early single digits only.

Reading the brand-switch matrix

The brand-switch matrix tracks two questions asked to the respondents of this survey. The first was, “Which brand do you currently own?” If they owned multiple brands, then the primary brand was recorded. The second question was, “If you were to buy or recommend the same product in the next six months, which brand would you buy/recommend?” All figures in the matrix are percentages of those who currently own the brand mentioned in the left. The diagonal represents brand loyalty. That is, those who own a given brand and said that they will buy/recommend the same brand. On the left is the current ownership; on the right, are future choices. Read along any row, and you will get the brand shifts for the brand given on the left. Read along any column, and you will get the shifts to the brand mentioned on top of that row. 

Anil Chopra, Krishna Kumar, Sujay V Sarma