How to get actionable insights by Analyzing Your Exit Pages with AI

This content is generated from the audio voiceover so it may contain errors.

(00:00) Welcome to all and to all to this new video of the canal of iSocialWeb in the go to explain today a form to integrate all the analyses of BigQuery with OpenAI that has a lot of functionalities and here bring you a small example the first obviously is to have the poised surgery in BigQuery in this case for example is a function that what analyses would be the ones of this table of data say that it is the one of the one of the account of Gpt 4 of social web from 1 March until yesterday Which are by users the pages that have had
(00:41) elder exit page that is to say that have had main start by the user No here is the listing of the surgery of the result costs And the interesting part is that once we execute okay can explore data And explore in the Notebook of Colab this what does Opens us a structure that already comes basically poised okay this of here is of here with the project the location and this part of here costs that it says us a bit that has the risk Scribd that gives us a bit the information that have we
(01:15) what add him is all the authentication of Open there is obviously install them all the necessary dependencies and afterwards what do Is to avert the problem of the maximum of tokens of the 4000 that has the Give Vinci 003 what do Is to work by Pareto and what say is Hears looks takes these starts calculates the total and afterwards Give me 20% of the urls that have 80% of the starts with which centre me in which have main impact and with this what do is once extract this information what do is enter it
(01:52) inside the prompt here that is to say are a list of data of downloaded the data of a web in which it shows the loss of the users by or rl that is to say the wireles by which the user has gone two points text analyses these data índica Which actions and in what urales would act to improve the conversion to improve the experience of the user costs also without 003 the maximum of token can change it and what #give is this #give the luck those that would act okay this is interesting because it gives us create
(02:21) that they are five in front of already can it see here 523 voucher afterwards gives us actions some are not specific obviously, but if no logical on the evaluation of the usability costs to improve the navigation restructure the information that is to say perhaps this part if we improve the user do not go because it does not understand at the end the exit does not know if it is due to that it has fulfilled And already and goes or to that really it is a satisfactory research with which was interesting Here also work in usability in improving the information
(03:00) but also go you here that it also works in the part to improve the navigation costs especially in this page imports the interesting here is that it is saying us Hears looks Perhaps to improve the success that is correct can follow improving the navigation Then this would be some proposals that would give us Open it is necessary is interesting for me that somehow what are Obviously with limitations of the prompt what are integrating is the surgeries of biquery that are data at the end is a Date frame what us
(03:30) it #give that it is the one who can see here no explains it to you Or was can not do that it interpret it is not very very very robust, but if it can give us ideas especially by where attack Hear Which are those that would attack or for example one another variant to this would be take all these URL Chooses the path only and tries categorizarlos the URL no because for example I that know this part of here no, as artificial Intelligence and artificial Intelligence As perhaps the group of Intelligence
(04:06) artificial has a lot of success in a lot of kit rate and, therefore, has to be worked in This Set No that would be the idea That is the Colab and at all expect that it was useful Thank you very much