Andrew Jenkins
Manabu Morita
Thanapol Poopunsri
Steven Stavrou
Siriorn Vichaiwatanapanich
Neoshop
Problem & Our Approach
Problem: Declining revenues and customer loyalty. Find
the best strategy to increase revenue from current
customers and retain them.
Our Approach: Analyzed data in SPSS and visualized in
Tableau to understand the difference of each loyalty
segment.
Limitations & Assumptions
• No gender information
• Only provided average spending
• Only one product (record) per customer
• No cost data
• Sales/Products could be seasonal
• Assume that conversion rate for a single channel is same for all
segments
• We don’t know physical store locations
• Based on current customer segments. Could be different.
Key Findings
Total Spending Avg spending/customerNumber of customer
• Silver has the most customers, smallest basket.
• Gold has the least customers, largest basket.
Key Findings
Key Findings
• Total Spending by product category
Key Findings
• Product sub-category by segment - Software and Book
Key Findings
• Product sub-category by segment - Video game
Marketing Campaign
• Run 3 outbound campaigns to get the
highest conversion rate through different
channels
• Direct mails on gold customers for two
weeks
• Run inbound marketing (Web & POS) to
support outbound messages (direct mail,
mobile apps, social network)
• Focus on specific product marketing for
Electronic (Gold and Silver) and Books
(Silver and Bronze)
Marketing Campaign
Segment Current
avg
spending
Additional avg
spending from 5
week campaign
% increase New Total avg
Spending
Gold 62.98 26.48 42% $59,941
Silver 11.29 2.92 26% $67,239
Bronze 19.75 6.61 33% $85,387
• Evolving industry
– Are the old segments still valid?
• Provided Limited data
– There may be other legitimate factors
Time to Re-segment?
Methodology
• Use XL Miner to transform variables
– Missing data
– Dummies
– 1 Bin
– Results in 21 segmentation variables
• Run the segmentation with ME XL
Output
Loyalty Test
Neoshop presentation-Marketing Analytics

Neoshop presentation-Marketing Analytics

  • 1.
    Andrew Jenkins Manabu Morita ThanapolPoopunsri Steven Stavrou Siriorn Vichaiwatanapanich Neoshop
  • 2.
    Problem & OurApproach Problem: Declining revenues and customer loyalty. Find the best strategy to increase revenue from current customers and retain them. Our Approach: Analyzed data in SPSS and visualized in Tableau to understand the difference of each loyalty segment.
  • 3.
    Limitations & Assumptions •No gender information • Only provided average spending • Only one product (record) per customer • No cost data • Sales/Products could be seasonal • Assume that conversion rate for a single channel is same for all segments • We don’t know physical store locations • Based on current customer segments. Could be different.
  • 4.
    Key Findings Total SpendingAvg spending/customerNumber of customer • Silver has the most customers, smallest basket. • Gold has the least customers, largest basket.
  • 5.
  • 6.
    Key Findings • TotalSpending by product category
  • 7.
    Key Findings • Productsub-category by segment - Software and Book
  • 8.
    Key Findings • Productsub-category by segment - Video game
  • 9.
    Marketing Campaign • Run3 outbound campaigns to get the highest conversion rate through different channels • Direct mails on gold customers for two weeks • Run inbound marketing (Web & POS) to support outbound messages (direct mail, mobile apps, social network) • Focus on specific product marketing for Electronic (Gold and Silver) and Books (Silver and Bronze)
  • 10.
    Marketing Campaign Segment Current avg spending Additionalavg spending from 5 week campaign % increase New Total avg Spending Gold 62.98 26.48 42% $59,941 Silver 11.29 2.92 26% $67,239 Bronze 19.75 6.61 33% $85,387
  • 11.
    • Evolving industry –Are the old segments still valid? • Provided Limited data – There may be other legitimate factors Time to Re-segment?
  • 12.
    Methodology • Use XLMiner to transform variables – Missing data – Dummies – 1 Bin – Results in 21 segmentation variables • Run the segmentation with ME XL
  • 13.
  • 14.