The study revolved around advertising for Gatorade in EA's games NHL 09, NHL 10, NBA Live 07, NBA Live 08, NBA Live 09 and NBA Street Homecourt, which featured the sugar water brand on places including arena signs, water bottles and score updates.
Households under Nielsen's watchful eye--over 100,000 of them--who purchased at least one of the games apparently spent an average of 24% more on Gatorade products than those which did not, the study discovered. According to EA, this offered a return on investment of $3.11 per household.
"Nielsen's study is a milestone for interactive entertainment," EA's senior VP of global media sales, Elizabeth Harz, said in the announcement. "For the first time, advertisers are able to link the value of their in-game marketing or sponsorship to actual sales. Now brands can feel confident adding gaming as a core media channel for their advertising."
In 2008, one of Nielsen's earlier studies into in-game advertising concluded it increased "positive brand attribute association" while a large majority of video game players--82% of those tested--felt games were "just as enjoyable with ads as without."
I wonder if they have metrics on how much the households that purchased sports games were spending on Gatorade before vs. after. Because, seriously, if their big finding is that IF you buy an EA sports game then you're statistically likely to spend 24% more on Gatorade than if you didn't buy that game, then I'd like to know how they're controlling for the possibility that people who buy sports video games, regardless of in-game ads, are always statistically more likely to buy Gatorade.
Here's what they have to say on methodology:
The study was based on Nielsen's US Homescan panel of more than 100,000 households, representative of the US population, including a subset of Homescan homes that scanned video game UPC (Universal Product Codes) barcodes. The scanned barcodes were matched to a reference library of more than 14,000 video game titles. Nielsen compared the households that purchased at least one of the studied games before and after Gatorade branding was integrated into the games (the test group) with households that didn't purchase one of the games (the control group).
These test and control group homes are projected out to the broad Homescan panel by matching them with the larger Homescan household universe based on similar purchase patterns and demographics in order to achieve a statistically reliable sample. Finally, the sales impact of Gatorade advertising was measured by analyzing and comparing Gatorade purchase behavior between the households that had and hadn't purchased the games that carried Gatorade advertising.
Seems iffy, the control group was self-selected, ie. their own personal preference on whether or not they chose to buy the game. That's dangerous as that variable might be linked to the one in the study, ie. gator-aide purchases.
People who buy sports games buy more gatoraide then people who don't buy sports games. The methodology has to be pretty exacting to actually glean some information on *why* that is the case (ie. because of advertisements in gatoraide).
It's why I never wanted to be a stats major ;)