Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

Marc Torrens @ Strands

1,097 views

Published on

Marc Torrens @ Strands

Published in: Data & Analytics
  • Be the first to comment

Marc Torrens @ Strands

  1. 1. ANALYSING BANKING DATA TO PROVIDE RELEVANT OFFERS TO CUSTOMERS Jim Shur, Chief Architect Ivan Tarradellas, Product Manager Marc Torrens, CIO © Strands Inc. 2014 Enric Plaza, Research Professor
  2. 2. WHAT DO WE DO? Strands is a global provider of personalisation and recommendation solutions to innovate in two sectors: financial institutions and online retailers. Strands Finance Strands Retail © Strands Inc. 2014 2
  3. 3. THREADS IN BANKING “Banks are losing their monopoly on banking.” – Francisco González, BBVA © Strands Inc. 2014 3
  4. 4. LEADER’S VIEWS “They all want to eat our lunch.” – Jamie Dimon, JP Morgan Chase CEO © Strands Inc. 2014 4
  5. 5. THE OPPORTUNITY “(…) the good news is that we still have one significant advantage, which is the vast array of financial and non-financial data that we accumulate.” Francisco González, BBVA CEO © Strands Inc. 2014 5
  6. 6. A NEW ROLE FOR BANKS Banks are faced with a great opportunity to move from being a mere provider of financial services to become a provider of solutions. © Strands Inc. 2014 6
  7. 7. TECHNOLOGY VISION BIG DATA MACHINE LEARNING USER EXPERIENCE Transactional Data > > Knowledge Actionable Insights Transforming Data into Knowledge to produce Actionable Insights for innovative Financial Software © Strands Inc. 2014 7
  8. 8. TURNING DATA INTO BUSINESS CUSTOMERS AND MERCHANTS VAST HISTORICAL DATA AVAILABLE NEW APPLICATIONS IN CONSUMERS’ DIGITAL LIFE © Strands Inc. 2014 8
  9. 9. How do I find the most relevant offers? HELP! CARD-LINKED OFFERS CLO How can I attract new customers? How well do I know Is my customers? © Strands Inc. 2014 9
  10. 10. HOW IT WORKS CARD-HOLDER’S PERSPECTIVE Accept Offer in Digital Banking Buy at Merchant Pay with Bank Card Get Cash Back in Your Account RETAILER’S PERSPECTIVE Get Charged from Bank Upload Offer in Digital Banking Monitor Offer Campaign Sell © Strands Inc. 2014 10
  11. 11. Hi, my name is Mario and I like wearing trendy clothes Hi, my name is Sarah and have a shop selling trendy shoes VIDEO © Strands Inc. 2014 11
  12. 12. THE PROBLEM I want RELEVANT offers I have MARKETING strategies Maximise the overall Performance for all offers © Strands Inc. 2014 12
  13. 13. THE PROBLEM I want I have MARKETING strategies Maximise © Strands Inc. 2014 13
  14. 14. RETAILER VIEW © Strands Inc. 2014 14
  15. 15. CAMPAIGN AUDIENCE FILTERS • A CAMPAIGN is defined by an OFFER and an AUDIENCE • An AUDIENCE defines a group of consumers by a set of filters: • Demographic filters (e.g. 20-30 years, married, in Barcelona) • Behavioural filters • Behavioural filters are based on Commercial Interests • Loyalty: how loyal is the customer to the merchant (loyal, shared, competitor) • Frequency: how frequent is the customer buying to the merchant (low, med, high) • Purchase Segmentation: the buying segment of the customer (low, med, high) • Location: where is the customer buying (in, out) © Strands Inc. 2014 15
  16. 16. MARKETING STRATEGIES © Strands Inc. 2014 16
  17. 17. MARKETING STRATEGIES © Strands Inc. 2014 17
  18. 18. MARKETING STRATEGIES SMART MARKETING STRATEGIES are an easy way to build audiences for different marketing goals, so merchants do not need to be expert marketeers and use complex filtering: SELECT STRATEGY Increase Loyalty Increase Spending Increase Frequency Win New Customers Reward Loyal Customers … SMART MARKETING STRATEGIES MERCHANTS REFINE (OPTIONAL) TARGETED AUDIENCE Increase Loyalty Win New Customers © Strands Inc. 2014 18
  19. 19. THE PROBLEM I want RELEVANT offers I have Maximise © Strands Inc. 2014 19
  20. 20. CARD-HOLDER VIEW © Strands Inc. 2014 20
  21. 21. USER CENTERED CAMPAIGN SALIENCE • The degree of interest of a campaign to a specific user: • Likelihood of buying in the category of the campaign • Proximity of the user to the merchant of the campaign • Activity of the user with the merchant that is making the campaign • Loyalty of the user to the merchant of the campaign • Merchant Fitness of the user considering the median of the merchant’s selling prices Likelihood (LK) Proximity (PX) Activity (ACT) Loyalty (LY) Merchant Fitness (MF) © Strands Inc. 2014 21
  22. 22. LIKELIHOOD OF BUYING IN A CATEGORY Supervised Learning Seasonality (12 Months) Customers (Millions) Categories (Hundreds) 2 YEARS OF DATA LAST YEAR MONTH TO PREDICT Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Training Set Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec h✓(x) Hypothesis Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec © Strands Inc. 2014 22
  23. 23. LIKELIHOOD OF BUYING IN A CATEGORY 12 Months Month Transactions Month Amount tree1 tree2 … treen k1 k2 kn voting k © Strands Inc. 2014 23
  24. 24. THE PROBLEM I want I have Maximise the overall Performance for all offers © Strands Inc. 2014 24
  25. 25. OVERALL SALIENCE CAMPAIGN SALIENCE indicates the priority of a campaign for the system. AR = campaign accomplishment ratio TR = time ratio gone for a campaign USER-CENTRED CAMPAIGN SALIENCE indicates the relevance of that campaign for the customer. combination of behavioural features + demographics % © Strands Inc. 2014 25
  26. 26. © Strands Inc. 2014 Comprehensive and interconnected set of solutions to leverage the value of customer data. 26 THANK YOU!

×