Transcript Slide 1
©2014 Experian Information Solutions, Inc. All rights reserved. Experian Confidential. Voice of Customer Analytics Julio Guedes Head of Analytics | Serasa Experian ©2014 Experian Information Solutions, Inc. All rights reserved. Experian and the marks used herein are service marks or registered trademarks of Experian Information Solutions, Inc. Other product and company names mentioned herein are the trademarks of their respective owners. No part of this copyrighted work may be reproduced, modified, or distributed in any form or manner without the prior written permission of Experian. Experian Confidential. Voice of the customer analytics We present an innovative project using the data obtained from the conversion of speech into text, associated with the application of text mining techniques Call center records Collections Converting voice to text Credit Fraud detection Ontologies CRM Analytics Marketing 3 ©2014 Experian Information Solutions, Inc. All rights reserved. Experian Confidential. Text mining approach Words net, tag cloud, and n-grams Applied statistical methodologies to identify relevant information in automatic text classification This debt is out of my family budget n-grams (trigram): debt_out_family out_family_budget Store and analyse textual data Storage and management of processed data 4 ©2014 Experian Information Solutions, Inc. All rights reserved. Experian Confidential. Voice interpreter – workflow Example of a call record Word identifier Filtering and frequency analysis Word classification ANN input 5 ©2014 Experian Information Solutions, Inc. All rights reserved. Experian Confidential. First case – Collections – The largest LATAM bank Itaú sent to us a sample of 1,602 calls from their credit card call center 111k transcribed terms 14k are distinct 2,260 positive words 1,877 negative words 6 ©2014 Experian Information Solutions, Inc. All rights reserved. Experian Confidential. First results – New classification Customer workplace – 14.4% Neighbor phone – 4.2% Relatives phone – 81.4% Will have money in the next months – 1% Effective negotiation – 97% Customer did a counter-proposal – 2% Unsuccessful contact – 5.2% Non-pay reasons – 8.8% Customer hung up the phone – 8% Debt belongs to a third party – 14% Improper time to contact – 5% No interest in paying the debt – 73% Family member with medical issues – 3.1% Unemployed – 90.3% Is already paying another debt – 5.2% Family problems – 1.4% We can identify who are the customers with the highest probability for paying their debts Information about the debt, the collection and payment: Recordings for the last six months Customer Operator Payment Debt value Due date Entry date 7 ©2014 Experian Information Solutions, Inc. All rights reserved. Experian Confidential. Commercial potential in the long run It is an application of Big Data Applied to any vertical That meets real needs – 100% of consulted clients are interested in Global Opportunity Potential unexplored 8 ©2014 Experian Information Solutions, Inc. All rights reserved. Experian Confidential. #FOIC2014 ©2014 Experian Information Solutions, Inc. All rights reserved. Experian Confidential. Julio Guedes Head of Analytics Serasa Experian e: [email protected] t: +55 11 2847 9238 m: +55 11 96381 8564 ©2014 Experian Information Solutions, Inc. All rights reserved. Experian Confidential.