Customer profiling: From natural language to web design. Presentation at CIRCLE, Estoril, Portugal in April 2010
1 of 10
Download to read offline
More Related Content
Presentation at CIRCLE conference Estoril
1. Customer profiling: From natural language to web designOuti-Maaria Palo-ojaPresentation at CIRCLE on April 9, 2010
2. 5.4.2010From natural language to web design / Outi-Maaria Palo-oja2eTourismProfilerFrom log data analysis to customer profiling on the InternetTo identify typical customer profiles and customer needs among the internet usersThis information is utilised in targeting marketing efforts according to customer profiles. The customer profiles are identified from the log file that is collected mainly from the travel web sites.ProblemPartnering organisations identified that customers still tend to phone and book their trip
3. Supplementary study on phone based profilingHow sales personnel identify customer needs and typical profiles?What are the critical events during the sales talk?Could we simulate these events on the internet?Is it possible to utilise the phone conversation based profiles on the Internet?How internet marketing is adapted according to customer profiles?5.4.2010From natural language to web design / Outi-Maaria Palo-oja3
5. Data and AnalysisDataCall center in travel industry26 conversations between travel agent and buyerVideo, screen capture, phone recordingAnalysisVideo annotation through VICTORWeb based multi-user video annotation tool Time sensitive comments5.4.2010From natural language to web design / Outi-Maaria Palo-oja5
7. Findings from the dataCustomers goals and preferences as a basisGroup 1: To book a trip under certain constraintsGroup 2: To make changes or seek additional informationGroup 3: To seek information for decision makingThe conversation between seller and buyerBrand new enquiries: DialogueSpecifications for bookings: Customer-leaded interviewNon-specific enquiries: Seller-leaded interview5.4.2010From natural language to web design / Outi-Maaria Palo-oja7
9. Entry page to divide users in threeWeb shopExactly expressed preferences (prerequisities & supplementary)Boolean operationsMore options through content-based recommendationService deskSign in with customer nroPersonally modified messagesExtra sales through collaborative-based recommendation Information deskor virtual brochureGeneral information with rich descriptionsUser generated content