Notes from GAUC2014 – Google Analytics Conference in Melbourne

Written by Mick Gibson on . Posted in Uncategorized

Spent an afternoon yesterday, topping up on analytics skills, with the LovesData team at the #GAUC2014 Google Analytics Conference.

For those of you not able to attend, here are all my notes (below) from the Masterclass session.

ALSO, see my workflowy notes from Friday’s full session with:

GA Conference 2014 Notes from Masterclass (link to notes on workflowy)

  • http://lovesdata.co/masterclass (download PPT presentation)
  • MASTERCLASS SESSION
    • Universal analytics
      • * Intro
        • Measure across devices
        • Measure online and offline
        • Customer centric
        • integrate with own data
      • Code:
        • _ga (cookie)
        • cid (client id)
          • Randomly generated
          • Date stamped
          • Sent with each hit
          • Will be different with each browser used
        • tid (Time id)
      • Manual firing of analytics:
        • http://www.google-analytics/collect?
          • v=1&
          • tid=UA-123456-1
          • cid=343423523&
          • t=pageview& (hit type eg page, event etc)
          • dp=%2Fexample&
          • z=2346557 (cache buster)
        • Details of all paramaters available at: http://goo.gl/OjVEZH
      • Measuring offline interactions
        • Examples of how/ where to use:
          • Point of sale
          • Event tickets
          • Coupons
          • Membership cards
          • Game consoles
          • Connected TVs
          • Movement
          • RFID
        • http://goo.gl/lKtqTf
      • * User ID
        • Ability to tie sessions together across devices/ touchpoints/ offline etc.
        • Can use multi-device reports (i.e. desktop > mobile > tablet etc)
        • Requirements:
          • Google’s requirements:
            • Full rights to upload data
            • Notify end users (consent or opt out)
            • Abide by other policies (eg remarketing)
            • No PII (Personally Identifiable Information)
          • Legal requirements:
            • Talk to a lawyer 🙂
        • USER ID Workflow:
          • 1. Review privacy and legal
          • 2. create new USer ID view
          • 3. Modify tracking
            • Javascript
            • GTM
          • Steps to create a user-ID (and separate reports)
            • Analytics > Property view >Tracking info > New reporting view
            • Analytics account structure
              • Property
                • Existing Voew
                • New View
                  • User ID
            • User ID Tracking code (in analytics):
              • {‘userId’:’744524′}); (ID string = dynamic)
            • Key to incentivise it so people do identify themselves (login, save settings, filters, wishlist, browse on another device etc)
            • DEMO
              • http://goo.gl/83qOxm (user ID groups under one customer/ user)
      • * Data Import
        • Why it can be useful:
          • ROI of advertising campaigns (cost data upload)
          • Merge CRM data
          • Add blog author details to posts
          • Customer Lifetime Value (CLV)
          • Demographics
          • Weather
        • Examples:
          • Cost Data for non-Google advertising
          • User Data about your customers (matching via userID)
          • Product data (such as brand or category)
          • Campaign Data such as sourceor medium
          • Refund data for ecomm transactions
          • Content Data such as author
        • User ID data import workflow:
          • Create a custom dimension with user scope
          • Joining data is key (user ID allows the combining of different data)
        • Cost Data Upload
          • Must be using campaign tags (source and medium are required)
          • Upload costs, impressions clicks
          • Demo:
            • Analytics > Admin > Property > Data Import > + NEw Data set > select (cost data)
            • Name data set.
            • There is a limit to number of data sets (25 data sources = max)
              • Can combine all data into a single data source (to save on data sets)
              • Or upload individually per channel (twitter/ facebook/ etc)
            • Select view
              • Link data:
                • Clicks
                • Cost
                • Impressions
                • Campaign (name)
              • Summation (default) Keeps a running total
              • Overwrite (overwrites all)
              • Get schema / downlaod CSV template (for upload purposes)
                • Date yyyymmdd eg 20120123
                • Cost analysis data report
                  • Can be customised (custom reports)
                    • ROI
                    • RPC
                    • Margin
  • GOOGLE TAG MANAGER – Johann de Boer
    • Tag management systems
      • Central place to control tags
      • Deploy new tags and changes quickly
      • Reduce reliance on IT to implement tags
    • Quick start guide to GTM setup
      • Create new account (preferably under your main google account)
      • New container (normally unique to website)
        • Website domain(s)
          • Http
          • Https
          • etc
        • Select timezone
        • Create + accept
        • Copy container code (add to website just below first <body> tag)
      • New Tag
    • Terminology and concepts
      • Tags
        • Code to be fired on pages
          • Remarketing
          • Conversion codes
          • Listener tags
            • Link click listener
            • Click listener
            • Form submit listener
          • Custom HTML tags
            • // insert coe
            • {{macros}} can included within custom HTML tags
          • Remember:
            • You cant use GTM for tags that need to be in the <head> of the page
            • Tags load asyschronously (can’t order firing)
            • Option for document.write is available
            • Dont use for A/B experiment tags
          • Data layer
            • A way for your website to consistently communicate with GTM
            • Data layer should become part of your website design, so it can scale with time.
            • History change – useful for AJAX apps
            • Timer – trigger a periodic event n time
            • Link click and form submit, uses call back.
      • Rules
        • What controls when a tag should fire
          • Uses information from macros to fire tags
          • A rule is one or more conditions which must ALL be true.
          • Conditional operators (and inverse operators):
            • equals
            • starts with
          • Firing rules (triggers the firing of a tag)
          • Blocking rules (stops/ prevents a tag from firing)
      • Macros
        • Placeholders for data
          • Javascript function macro
            • // Calculate tax
            • function() { return {{total value}} * ({{tax rate perc}} / 100)}
    • How to control tags and connect data
    • Best Practices
      • Put GTM on every page
      • Design and implement a data layer early
      • Keep your tags and macros organised (naming convention etc
      • Use macros to avoid repetition (eg. create macro for web property ID)
      • Preview changes across browsers
      • Name versions
      • Limit publish access to trusted people only.
      • Debug before publish
      • www.lovesdata.co/new-benchmarking
  • GOOGLE ANALYTICS APIs
    • Data Collection APIs
      • Collect and send data to Google Analytics for storage and processing
        • Website tracking
        • Mobile app tracking
        • Offline server-side measurement (eg via measurement protocol)
    • Configuration APIs
      • Configure Google analytics accounts (programatically for larger businesses in need of automation)
        • Import offline data
          • Ecommerce refunds
          • Cost data etc
        • Get and set configuration details
          • List accounts, properties views
          • Goals, filters and view settings
          • User permissions, segment definitions
        • Provisioning of new accounts
    • Management APIs
      • Collections of resource entities, including
        • Accounts properties, views
        • goals experiments
        • Linked AdWords accounts
        • Data uploads
        • Un-sampled reports (premium)
      • Read write access
    • Reporting APIs
    • Embed API: https://developers.google.com/analytics/devguides/reporting/embed/v1/devguide
      • Embed reporting widgets into a webpage
      • Build an interactive dshbaord
      • Simple authentication process
      • lovesdata.co/UH1WK (code samples)
      • Embeddable components
        • Authentication button
        • View and Date range selectors
      • Implementing:
        • Include the embed API javascript library on page
        • <section> tags to contain each component
          • Authentication button
          • View and date range selector
          • Data visualisation
        • Javascript to integrate components and handle events
      • Build your own applications:
        • Setup a Google APIs project
        • Get familiar with documentation
        • Experiment with queries
        • Use client libraries already set-up by Google
  • KLIPFOLIO – Michael Fridman
    • Stop data ‘pukes’ .. less is more.
    • Good visualisation:
      • Pie-charts
      • Simple, accurate, compelling and tell a story
    • Bad visualisations:
    • Great tool for designing good dashboards: lovesdata.co/abela-chart
    • When is Klipfolio useful:
      • When analytics is not enough
      • Other/ multiple data sources
      • Want to customise.
    • Model own data dynamically in Google docs
    • Summary:
      • Great for automating and integrating data
      • Query directly/ indirectly
      • Choose a chart that’s appropriate to data and message
      • Multiple data sources
      • Custom visualisations
      • Klipfolio.com/resources
      • youtube.com/user/klipfolio

Comments (1)

  • Brucey

    |

    Thanks for sharing! And I have to agree with you, “no need to attend a ‘Masterclass’ to learn this” — that half day would have been better spent reading up on the subjects presented. Why did they assume that masterclass attendees wouldn’t already know all this stuff? An extremely expensive day and a half without one solid takeaway I can use. Bah!

    Reply

Leave a comment