beverages

The In-Depth Research Behind Pepsi Refresh 2011

Singh

In late 2010, PepsiCo announced a number of changes for 2011 to its crowd-sourced community donations program, the Pepsi Refresh Project.

Among these: Offering twice as many grants but putting more emphasis on smaller denominations; honing the categories in which consumers can submit project ideas down to four (arts and music, communities, education and a rotating "Pepsi Challenge" category); adjusting the balloting time frame and procedures; and enabling people to vote through a wide range of social media, including Facebook, Twitter, the Project's mobile app and the text2vote program, as well as through the Refresh site.

But exactly how did PepsiCo come to its decisions to make these and other tweaks to the Refresh program for its second year of operation? The research processes, recently shared with Marketing Daily by Shiv Singh, PepsiCo Beverages head of digital*, offer an eye-opening look at the depth and sophistication now possible through analysis of social media and other digital consumer feedback.

advertisement

advertisement

PepsiCo's research started last August/September, with in-depth surveys of the brand's Facebook fans. Each week, for four weeks, PepsiCo asked fans a different set of predetermined questions -- each set organized around a different aspect of the program and designed to elicit feedback on what these consumers felt had worked and not worked during the project's first year.

For example, questions probed areas such as how fans voted during 2010, what community idea categories were most meaningful to them, what they liked/didn't like about the comments capabilities, and how they went about spreading the word/encouraging votes for ideas that they had submitted.

PepsiRefresh-B

"We basically took the framework of focus groups and moved it online," says Singh. "Unlike a traditional survey questionnaire, respondents could see how others had responded, for example." PepsiCo analyzed each week's responses in a variety of ways -- and put somewhat more weight on early responses, to take into account the influence factor always present in focus group-like settings, he explains.

At the end of each week, PepsiCo shared with its Facebook fans a summary of the results -- meaning the major themes expressed by the fans that week and how PepsiCo would likely adjust the program in 2011 to reflect his input.

The next stage involved partnering with a social media research company to deeply mine the content and implications of social media conversations across all channels. "We wanted to understand when and where people talk about the Refresh project -- the frequency, patterns and velocity of conversations -- and what specific topics triggered emotions and conversations, and why," says Singh.

For example, PepsiCo at this stage had a handle on which idea categories mattered most to people, but wanted both to confirm those and probe exactly what about each of these categories resonated most with people -- in part to inform the rotating "Pepsi Challenges" that it would be introducing in 2011.

"With music, for example, we found that music's ability to make people feel good mattered much more to them than the artists themselves or the media or channels they used to find or share the music," says Singh.

This stage also revealed that outside of the Refresh site, YouTube was by far the biggest hub for generating conversations about the project. "People's responses to the videos generated off-the-charts talk levels, leading us to incorporate video in our revamped Refresh site to a much greater degree," reports Singh. The insights also led to a more "people-centric" approach on the site, including more in-depth looks into the people and projects awarded, he adds.

The next stage included in-depth analysis of all 120,000+ ideas that had been submitted in 2010, to identify the hot issues and how people feel they can make a difference in America by helping their communities -- as well as an analysis of who submitted and who won awards. (As has been reported, the reasons for adjusting the grant levels and voting rules had to do with the finding that large organizations, including health interest and religious groups, had been dominating the awards by leveraging alliances to ensure large popular votes for their causes.)

Next, PepsiCo's digital team analyzed all of the comments on the Refresh site to date -- 20,000 per day on average, according to Singh. (These site-comment analyses, along with monitoring and analyses of social media conversations relating to Refresh, are ongoing.)

In the last research stage for the 2011 Refresh changes, PepsiCo went out and did in-depth, face-to-face interviews with 2010's award winners.

"It's like peeling an onion -- you gain a deeper and deeper understanding with each type of research," sums up Singh.

*The article was amended.

Next story loading loading..