Solution providers were invited in late March and April 2007 via e-mail and online to fill out an online ballot rating 2006’s products by category. The online ballot first asked voters their involvement in each category; then, among those categories in which the voter resells products, five were chosen at random for detailed ratings of products.
Ballots were de-duped based on user IP address and e-mail address, and non-channel respondents’ votes identified and suppressed, to nullify attempts at ballot stuffing. In addition, suspicious voting patterns were identified and their associated ballots also suppressed. 350 qualified ballots were counted in all.
Ratings were provided by respondents in four areas for each product: familiarity with product, profit potential of the product to the channel, opportunity the product presents for associated sales, and overall market impact of the product. All ratings were based on four-point scales except for market impact, which was a three-point scale.
A voter’s ratings in all four areas for a product were combined to create a single voter score, from 0 to 10, for that product.
Profit potential counted for 4 of the 10 points, opportunity for 3, and impact and familiarity each for 1-1/2.
The voter’s familiarity level was also used for weighting, so that the greater a voter’s familiarity with a product, the more that voter’s ratings of the product would count.
If any rating was skipped for a product, the entire vote for that product was considered spoiled, and no voter score created.
Once all voter scores were calculated, the average of all available scores for each product was determined, giving that product its final score.