Retail social media segmentation

Retail social media segmentation

• 1.
  © 2013 IBMCorporation Social Media Segmentation Use Case Sales and Delivery • Utilize publicly available social media and other external data sources to determine the digital social behavior of the U.S. Millennial population ages 18-30 yrs. • Leverage behavioral insights of digital social expressed preferences to build Millennial Social Segments relevant to the Retailer in order to better engage Millennials • Provide a methodology and set of technologies to operationalize the listening and context gathering insights for the Retailer into a repeatable process Project Objectives IBM Sales & Distribution Solution Sales Business Solution Professional Tel. +49 151 1252 1320 [email protected] Oliver Seidel Foto
• 2.
  © 2013 IBMCorporation Today’s consumers have little interest in retailers who don’t meet their expectations Anticipate and service my shopping needs Use what you know about me to deliver a seamless and interconnected engagement regardless of where I am shopping Make it convenient and easy to interact with you Consistently capture my history and profile and deliver information and services to me through my preferred channel Be there when I need you, in real time Empower me with real-time updates, proactive communications and choices before, during and after I shop Remember all of our interactions Understand me and my preferences through our past interactions 1 Retail Industry Challenge Retailers must engage in “social business” and use social data in order to acquire advocates and eventually, evangelists for the brand Leverage social content to create meaningful, productive, two-way dialogue…It’s a conversation! Use insights generated from those conversations to facilitate change across the retail enterprise Keep the relationship alive. Listening often and intently will create lifelong customers Let consumers know they’ve been heard and give them content to share
• 3.
  © 2013 IBMCorporation Solution Description – What is SMA (Social Media Analytics)? Reporting and Dashboard capabilities, Cognos v 10, single log-in for SMA project, with under-the-hood configured big-data feeds from social media through BoardReader. Includes SaaS option to complement continued momentum of on-premise deployments Through SMA, algorithms, conducting sentiment analysis, which provide output as to which segments have negative or positive comments or responses to a particular topic, product, or query. Segmentation from sentiment analysis can be conducted in a meaningful way. Another feature is that of evolving topics. Through a topic detection algorithm – after certain words and topics have been selected, other related topics will continue to be identified keeping the model fresh and relevant for as many days as selected. Importantly, because the dashboards are easily navigable, SMA is appropriate for many users within an organization from the BI user to the more technically advanced. SMA supports over 35 languages through BoardReader (Effysis), including Chinese Modern and Traditional. 2 SMA is a powerful tool that combines web-scraping of millions of social media feeds, sentiment analysis, segmentation,and search for evolving topics, in an intelligent reporting and dashboard interface. It supports over 35 languages. IBM Solution Capabilities & Use Case
• 4.
  © 2013 IBMCorporation Capture Analyze/Predict/Target Execute & Empower Advanced Customer Analytics StatisticsText Mining Data Mining Data Collection delivers an accurate view of customer transactions attitudes, opinions, and preferences Advanced Analytics capabilities enable retail brands to more effectively target customers for interactions; increasing response rates, improving results, enhancing the customer experience Targeted and customized interactions are delivered through appropriate channels at the brand’s discretion or on-demand Optimization Robust Data Collection Highly Customized Customer Interactions Solution Diagram – IBM’s Approach Capture customer data, Analyze the data to uncover insights and Execute on those insights to deliver quantifiable business results and improve the customer experience 3 IBM Solution Capabilities & Use Case
• 5.
  © 2013 IBMCorporation Millennial facets are modeled into Millennial themes adding context, depth and breadth to the analysis 4 Millennial themes (Feature Vectors) are derived in three ways to ensure the depth and breadth of the analysis: ACT – Which characteristic did the author discuss the most? N_ACT – How many different characteristics does the author mention within the facet? SUM – how many total conversations were captured about each facet? Feature Vector Retailers_ACT Retailer 1 N_Retailers 4 SUM_Retailers 75 25 core Millennial facets 163 key feature vectors (themes) modeled Selected 12 Clustering vectors (themes) • Education Attained • Value Shopper_ACT • Value Shopper • Young Millennial • Older Millennial • Max Topic Engaged • Fashion Style • Retailers_ACT • N_Retailers • Leisure Activities • Hobbies Engaged • Shopping Lifestyle MillennialTheme Creation 2.03M Authors Retailers Author # Retailer 1 Retailer 2 Retailer 3 Retailer 4 Retailer 5 18903 40 12 8 15 0 IBM Solution Capabilities & Use Case
• 6.
  © 2013 IBMCorporation Millennial themes give way to the 12 building blocks used to assemble the Millennial Social Segments 5 The 7 Millennial Social Segments are: • Highly homogeneous – creates segments with similar narratives • Highly differentiated – ensures distance between segments • Highly actionable – facilitates highly- specific targeting The 12 clustering variables were chosen through analysis of: • Number of conversations • Distribution within each millennial theme • Low correlation between millennial themes • Most importantly business understanding and implications Education Attained Young Millennial Older Millennial Max Topic Engaged Fashion Style Leisure Activities Hobbies Engaged Shopping Lifestyle Retailers_ACT N_Retailers Value Shopper_ACT Value Shopper Millennial themes and source are used as profiling variables to enhance each Millennial Social Segment Gender Gender Cluster # Male Female Macy's Enthusiast 15.91% 36.74% Lady Candid 8.55% 60.86% Savvy Shopper 14.22% 41.57% Coupon Lover 12.15% 47.66% The Everywhere Butterfly 12.54% 43.59% Fashion on a Dime 14.07% 43.27% High Fashion Connoisseur 11.74% 36.99% Brand IBM Solution Capabilities & Use Case
• 7.
  © 2013 IBMCorporation Millennial Social Segments 6 Brand Enthusiast Retailers and fashion style dominates their conversations Retailer 1 is discussed the most Share most via Twitter 49% of the population Value conscious; lives on a budget Finds and shares great deals Finds great value at Retailers 4 & 5 18% of the population Savvy Shopper Cluster 1 Cluster 3 Vocal multi-style Girly Girl Always out and about doing what she likes Finds what she likes at Retailer 2 and Retailer 1 8% of the population Cluster 5 Everywhere Butterfly Share! Share! Share! Blogs about everything fashion Sophisticated and preppy styles 1% of the population Lady Candid Cluster 2 Constantly looking for the best deals Vocal about everyday needs and style Hobbies reflect what she values most 0.3% of the population Coupon Lover Cluster 4 bloggers Engages in sporty and active lifestyle Young professional value shopper Fashion style matches sporty lifestyle 13% of the population Fashion on a Dime Cluster 6 High Fashion Connoisseur High fashion and sophisticated styles Prefers high-end retailers Jetsetter to the core 11% of the population Cluster 7 IBM Solution Capabilities & Use Case