23. “every 100 ms (thatʼs 1 tenth of a second) increase in page load time of Amazon.comdecreased sales by 1%”
24. 2. Matching products to customers @nilanp #BDLDN
25. @nilanp #BDLDN
26. Size ?
27. How do you match Size ?customers to products ?
28. Recommendation @nilanp #BDLDN
30. Recommendation @nilanp #BDLDN
31. Segmentation @nilanp #BDLDN
32. Data science
33. Segmentation @nilanp #BDLDN
35. Size ? @nilanp #BDLDN
38. Personalisation @nilanp #BDLDN
39. Carregistration Product type Customer flikr house
40. Decision trees @nilanp #BDLDN
41. Bayesianstatistics @nilanp #BDLDN
42. Graphdatabases @nilanp #BDLDN
43. Carregistration Credit Card Customer Email Address house
44. 3. Conversion
45. 3 millioncustomers @nilanp #BDLDN
46. 30 million visitors @nilanp #BDLDN
47. 200 million clicks @nilanp #BDLDN
48. Why are wedown / up?
49. GA picture
50. 0 USER JOURNEY! Only 6.3% of customers who enter the funnel from airport hotel pages end up converting! Holiday Extras domain! Commentary! Performance! Availability page! 1 in 7 visitors proceed to the upgrades page! ! lack of information?! 15%! ! lack of urgency?! Upgrades! Almost all proceed to the payment page! 98%! ! Performing well ! ! Can additional upgrades be presented to increase order value which may offset the impact of conversion rates?! Payment! 43%! Just under 1 in 2 then proceed to make a payment! ! Suggests the payment form is performing very well! Conﬁrmation! Summary! At present only 6.3% of users who enter the funnel make a purchase. The Total funnel efﬁciency! largest hurdle is on the availability page where only 15% make it through to the 6.3%! next phase.! We therefore focus our recommendations on the availability page in order to prevent customers falling out of the funnel at the availability page, there are a number of solutions:! a) Provide user with sufﬁcient information to assist in the decision process ! b) Provide user with clear steps to purchase so that they are able to clearly understand where they are in the funnel!
51. Why ? @nilanp #BDLDN
52. Causality ! @nilanp #BDLDN
53. Qualatitive data +Quantiative data = Causality @nilanp #BDLDN
54. Exit survey
55. nformation!e more Sample feedback!rmation.! Images! 13%! “Cant see a picture of the hotel. I like to see what I am getting not just have thewas more symbols for the ameneties”! Descriptions and Info! 65%!roduct “cant ﬁnd enough info about the undercover hotel”! Extras! 21%!down by “for early ﬂights, would be useful to see times of transfers. also info as to how to get to hotel if travelling by train.”!
56. ✦recapData Recap.... science
57. Capture Slice data dataLaunch Innovate @nilanp #BDLDN
58. ✦recap A 3 step action planData TODAY !start science you should
59. 1. Recruit data scientists 2. Store everything3. Start playing with your data @nilanp #BDLDN