A Dynamic Duo: Survey & Social Media Analytics

Donna Smith

Editor’s note: This guest post is written by Donna Smith, a professor at the Ted Rogers School of Retail Management at Ryerson University in Toronto, Ontario. Professor Smith uses Sysomos as a part of our ‘Sysomos In The Classroom’ program. This is a two-part series around research Professor Smith is conducting alongside her colleagues using Sysomos MAP.

“Do you own a pair of jeans?” This simple question led my colleagues and I down a multitude of paths to gain insight about a product that many of you may be wearing today.

As academic researchers and industry analysts, we bring you part one of our story on how an analysis and survey of social media data using Sysomos MAP and GAZE can help you understand what leads a consumer down the path to purchase.

We are searching for the key to how social network marketing impacts the purchase of jeans. The Holy Grail is to utilize consumer insight to increase ROI for jean brand manufacturers and retailers.

Why jeans? As a universal product, we knew consumers would have diverse views on the jeans in their closet. If we could gain insight into how social network marketing may be used to increase the sale of jeans, this knowledge could be applied to understanding the purchase decision drivers behind other products in apparel and footwear, a sector which has experienced a high rate of growth in e-commerce sales in North America.

Highlights of Our Survey

We began by surveying nearly 800 Canadian consumers on eight brands of jeans: Buffalo, Citizens of Humanity, Denver Haye’s, J Brand, Joe Fresh, Lee, Levi’s, and Old Navy.

Our data set had a 60/40 female/male distribution with 42% in the 21-40 age group. Of the individuals surveyed, 61% were on Facebook, 32% were on YouTube, 25% were on Twitter, 23% were on Pinterest, 23% were on Google+, 20% were on Instagram, and 3% were on Vine.

In order to predict purchase intent, we developed a model that would help us gain consumer insight; we examined the effect of pre-purchase influences, online brand engagement, fashion brand involvement (how the consumer develops an attachment to the brand), and how these three may be used to predict purchase intent in the context of social network marketing.

Survey questions probed the ways in which specific social media platforms, commercial users, or friends or family influence the consumer. For example, “Please rate how much you agree or disagree with: Retailer tweets influence my buying behavior.” We examined purchase intent from multiple perspectives, including hard (i.e., price, delivery) and soft (i.e., trust brand, consumer buzz) factors.

We then performed a segmentation analysis. Two segments emerged. The first was a “fashionista,” whose route to purchase was led through user blogs and fashion brand involvement. To immerse him or herself in the brand, this segment avidly searched for information using Twitter, Pinterest, and retailer websites.

Our findings: To appeal to this segment, social media marketers should push out content that keep this consumer au courant of the latest fashion trends relevant to the product or service.

In contrast, the second segment is the “social shopper.” His or her route to purchase is through word-of-mouth, user blogs, and online brand engagement. This consumer is very conscious of fitting in socially and wearing the “right” brand.

Our findings: In order to lead consumers in segment two to purchase, social media marketers may consider the development of campaigns that are authentic, in order to establish trust.

Overall, our segmentation approach demonstrated that there is no “one-size-fits all” for understanding purchase intent on social network sites, for consumers in this research.

Next Steps with Sysomos

Our work is far from done! We are now analyzing which messages and images would be appropriate for each segment, according to the brands purchased. Preliminary research using Sysomos MAP on three of the top-selling jeans brands in North America (Levi’s, Old Navy, and Lee) demonstrated that influencers with authority scores above 8 on Twitter belonged to commercial Twitter handles, such as online stores.

Tweets consisted of promotional messages for specific products and in some cases, influencers overlapped across brands.

In part two, we’ll dig deeper to analyze messages of influencers on different platforms and see if they correspond to the findings in the segmentation analysis. We will also analyze images using Sysomos Gaze in order to see if they coincide with our segment findings.

Additionally, we will determine what next steps would be most effective in moving consumers to purchase intent. We also hope to test our model in different countries and contexts to see if jeans are indeed a universal product.

twitter brand communities

 

wordcloudAs researchers, we see the benefit of gaining consumer insight through using multiple tools. This helps social media marketers plan campaigns where messages and images are up to date, meaningful, and nuanced.

Researchers:
Donna Smith
Ted Rogers School of Retail Management
Ryerson University
Toronto, Ontario

Tasmina Afroze
Student, Master of Science in Management
Ted Rogers School of Retail Management
Ryerson University
Toronto, Ontario

Ángel Hernández-García
Universidad Politécnica de Madrid
Madrid, Spain

Ángel F. Agudo Peregrina
Universidad Politécnica de Madrid
Madrid, Spain

Joseph F. Hair, Jr.
Kennesaw State University
Kennesaw, Georgia

Leave a Reply