Your ultimate guide to Google Analytics Remarketing.
Find out how Google Analytics is working and what are GA Audiences. Understand which signals to use, how to create Audience lists and and leverage GA capabilities for advanced uses cases such as cross-device Remarketing, CRM-based Audience Lists, Email-based Retargeting, etc ...
2. Intro.
Luigi Reggiani
Head of Data Analytics, Google EMEA
@LuigiReggiani
Thanks to all friends & colleagues that have helped me in creating this guide,
from GA Product teams to Product & Sales Specialists in Google EMEA.
Made in Paris - November, 2015.
3. Contents.
• Launch Presentation Mode & Pick a section from below •
AN INTRO TO
GA RMKT
GOOGLE
ANALYTICS
The Basics
Audiences
in Google
Analytics
BUILDING
AUDIENCE
LISTS
UNDER
DA HOOD
‹ Click this icon anytime to get back here.
LISTS INDEX
& RESOURCES
7. Simpler, Faster, Better.
1 TagSAME ONE FOR ALL PAGES
250+DIMENSIONS & METRICS
UNIQUE
FEATURES
GA yields Ultra-Precise Remarketing
Turning Moments that matter into Audiences to Consider
+ +
Across Google Search* & the Google Display Network
* Some GA Signals are not available for Remarketing Lists for Search Ads (RLSA)
--
8. GA allows to go deeper.
A 25-34 AGED VISITOR VIEWED
PRODUCT X PAGE 5 TIMES
ON PAST WEEK
THIS LOYAL PURCHASER, ASKED FOR
A TOTAL REFUND AFTER 1 WEEK.
HE IS A YOUTUBE SUBSCRIBER AND LIVES IN PARIS.
THIS WOMAN, FAN OF TRAVEL HAS
REMOVED 1 PRODUCT FROM HER CART
WHICH CONTAINED 2 PRODUCTS.
LAST VISIT FROM THIS
GOLD CUSTOMER WAS 1
MONTH AGO
THIS iPHONE 6
USER BOUGHT FOR
100$
IN 1 YEAR
THIS USER DISCOVERED OUR
BRAND THROUGH FACEBOOK
--
9. in-App also gets better.
The power and precision of Google Analytics
& powerful Google demographics data
to act on your audiences and re-engage with them.
USE GA AUDIENCES FOR :
IN-APP MONETIZATION IN-APPS REMARKETING
--
10. GA+RLSA* Case Study
CVR +65% AOV +58%
CPA -50%
• Advertiser : TransUnion, leading financial services provider
• Objectives : Find more customers, reduce waste in Search campaigns
Strategy :
○ Splitting the audience between new and existing customers.
○ Bid higher on Google search ads for new customers.
○ Spend less on converted customers.
Results :
- Full Case study available on Google Analytics Blog -
--
* Remarketing Lists for Search Ads
11. GA+GDN* Case Study
AOV +13%
1300% ROI with PRECISION REMARKETING
• Advertiser : WatchFinder, leading U.K. retailer of premium, pre-owned watches.
• Objectives : Encourage site visitors who don’t purchase initially to return and
complete a purchase & Encourage footfall to new London boutique.
Strategy :
○ Created 20 highly focused lists of visitors who demonstrated intent
○ Defined groups according to funnel stage, location, language,
on-site behavior, ISP and more
Results :
- Full Case study available on Think Insights Website -
CPA -34%
--
* Google Display Network
13. Data Collection.
For every page containing that code, GA Servers record a
Pageview Hit
with a Client iD attached.
GA SDK
for Apps.
Measurement Protocol
works for almost
everything else
Android & iOS
It works
for WEB & mWEB.
GA.js
(Legacy GA tag)
DC.js
(Legacy GA RMKT tag)
Analytics.js
(Universal Analytics tag)
--
14. Data Processing.
ONCE COLLECTED,
Data is then Processed
IN VARIOUS TABLES.
Originating device is Geolocalized, Categorized ...
HITS ARE SESSIONIZED, FILTERED, AGGREGATED, COMPRESSED, Etc, Etc ...
(you really don’t want to know all of it, don’t you?)
Finally, the magic happens, Data becomes
AVAILABLE FOR REPORTING
--
15. Data Reporting.
Accessible 24/7
through GA UI & GA APIs
Intra-day
FOR SMALL
PROPERTIES
The Day After
FOR BIG PROPERTIES
(Non Premium)
In less than 4h
FOR BIG PROPERTIES
(with Premium)
--
Real-time Reports
Location, Events, Traffic sources, Conversions, Page-views, Active Users
+
16. What Else ?
Event Tracking
FOR EVERYTHING WHICH IS NOT A PAGEVIEW
Videos
Plays
Cart
Addition Forms
Progress Documents
Downloaded Email
Openings
Tracking the above needs
ADDITIONAL CODE ON SITE
But it can be done in few minutes only through GTM,
--
17. Event Tracking in 30s.
For example : MEASURING HOW USERS INTERACT WITH VIDEOS
ga('send', 'event', 'Video', 'Click', 'Play', 4);
Required to send
Event data to GA
Event
Category
Event
Action
Event
Label
Event
Value
OPTIONAL
--
18. Social interactions.
For Example : MEASURING HOW USERS SHARE YOUR CONTENTS
ga('send', 'social', 'facebook', 'like', 'targetUrl');
Required to send
Social data to GA
Social
Network
Social
Action
Event
Label
--
19. Enhanced eCommerce.
--
Increase your Visibility
ON YOUR CUSTOMERS JOURNEYS
Product
Views
Cart
Addition &
Abandonment Transactions
& Refunds Internal
Promotions Checkout
Abandonment
Tracking the above needs
ADDITIONAL CODE ON SITE
20. Custom Dimensions & Metrics.
CRM iD
Customer Type
Membership
LTV Group
First Purchase Date
UP TO 20*
Custom
Dimensions & Metrics
# of Purchases
Offline Revenue
Ongoing value
Age
Temperature
* 200 with GA Premium
--
21. CRM
DATA
Sending Data to GA.
Offline
DATA
Offline
Transactions
Customer
Status
Offline
Events
Customer
Lifetime value
--
22. Data Security & Privacy.
Google classifies Google Analytics data as
CONFIDENTIAL INFORMATION
#NOT PIIs DON’T MESS
WITH OPT OUT
FROM COLLECTION TO REPORTING,
DATA IS KEPT SECURE & GUARDED AGAINST ANY THREAT.
#GET USER
CONSENT
--
24. Show me the Cookie.
User
visits a Site
for the 1rst time
A GA Cookie is Dropped
in his browser, on this Device
This Cookie contains
a unique Random
ID number
_ga=1.2.3456.77777
Important note :
GA only drops
a GA Cookie.
UA-999999-8
--
25. Cookie Matching.
_ga=1.2.3456.77777
When a User visits a site equipped with GA :
GA checks if there’s also an Advertising cookie
in the same browser
If Yes
Then a Cookie Matching is done, and both iDs are
attached in our servers for a limited period of time.
UA-999999-8
If No :
User can’t be retargeted
iD=28DDZLK987DHDKZ
.doubleclick.net
Note : RLSA uses a different Advertising Cookie that the one above. On Apps, GA RMKT relies on AdID or IDFA & on GA Apps client iD.
--
26. GA Audience Lists.
Use GA Remarketing to Define Audience lists using one or many of
250+ Dimensions & Metrics
--
*Not available for RLSA
BEHAVIOR
Sessions
Session duration
Bounces
Page depth
New Users
Returning Users
...
EVENTS
Conversions
Downloads
Video played
Social actions
Forms compl.
...
ECOMMERCE
Transactions
Revenue
Products
Add to cart
Remove from cart
...
ACQUISITION
Campaigns
Sources
Keyword
Ad Creative
Referral
...
DEMOGRAPHICS
Age group*
Gender*
Interests*
Affinity cat.*
Location
language
….
TECHNOLOGY
Device
Device Model
Browser
OS
...
27. GA Audience Lists.
As soon as a (retargetable) User matches list criterias,
his Advertising cookie iD is added to the Remarketing list.
+ +
Use GA Remarketing to target users doing multiple actions during a single session (visit)
eg. : Users that visited more than 2 pages during a Visit
iD=28DDZLK987DHDKZ
.doubleclick.net
--
28. GA Audience Lists.
+ +
Use GA Remarketing to target users doing multiple actions during multiple sessions.
eg. : Users that visited more than 2 pages
iD=989DDZIZ87DHDKZ
.doubleclick.net
Visit 2Visit 1
GA offers this unique ability to track activity
over time during periods of 7/14/30 days.
› This is called the Lookback Window
If a (retargetable) user match lists criterias during that
time frame, his Advertising cookie iD will be added to
the Remarketing List
--
29. The Look-back window.
Day 1
Purchase
Value : 30
List Definition : Users that generate more than 100 of Revenue
Look-back window : 30 Days | Membership duration : 180 Days
Day 20
Purchase
Value : 30
Day 40
Purchase
Value : 80
Total Value in 30 days = 60 : Nothing happens
Total Value in 30 days = 110 : User matches List Criteria. He is added in the Audience List.
User remain in the Audience list for 180 days
Day 221
User is dropped
from Audience List
--
iD=989DDZIZ87DHDKZ
.doubleclick.net
30. Smartlists.
Let Google Analytics Machine Learning automatically target best Users with Smartlists.
Smartlists can be based on your own history if enough data or Industry Benchmarks if not.
iD=3029K987VCDKL
.doubleclick.net
Machine learning uses dozens of signals, including location,
device, browser, referrer, session duration, and page depth
to identify the users who are most likely to
convert in a subsequent session.
--
31. Pre-Population.
UP TO 10 DAYS FOR DISPLAY
UP TO 6 HOURS FOR SEARCH
Google Analytics can pre-populate your Audience lists with up to 10 days of data
Only available if Data Collection has been enabled before that time frame.
--
32. User Buckets.
Create compositionally identical Remarketing Audiences for testing.
By default, GA assign each user into a User bucket with a number between 1 & 100.
Users Bucket : 1-25
Test Group with Bid Modifiers on Audience
Users Bucket : 60-70
Control Group, with standard Bids
--
You can create two or more compositionally identical Remarketing
Audiences and run a different ad campaign for each audience.
You can then compare the performance of each campaign and
determine which is more effective.
34. Once you Enter in a List, you stay in that Remarketing List...
… till List Membership duration is expired.
Good to know.
Once you Enter a Remarketing List, you can be retargeted less than 1 minute after.
GA RMKT Works in REAL-TIME.
All Lists created from GA can be re-used in AdWords to create
Custom Combinations & Similar Users
--
35. There’s More ...
--
DYNAMIC REMARKETING, with GA
Re-engage users on the Google Display Network,
based on the content they’ve already seen on your site.
37. Enabling Remarketing.
Step 1 : Make Sure AdWords & GA are linked
Step 2 : Enable Data Collection
Toggle Remarketing button in : Admin › Tracking info › Data Collection
Privacy Law applies in EU : User consent to collect ‘any data used for Remarketing’ is mandatory in most countries
--
38. Creating an Audience List (1)
Step 3 : Create a new Audience
Click [New Audience] button in : Admin › Property › Remarketing Audiences
--
39. Creating an Audience List (2)
Step 4 : Choose the GA view (where data comes from) and the AdWords account (where Data will go)
Select the AdWords account
Select the relevant GA View
--
40. Creating an Audience List (3)
Step 6 : Choose a predefined Template, create your Own List, . or import a Segment
To create your own
list using the
Audience Builder
click the
Create New
Button.
We have templates for
most common lists :
Recommended Audiences
Pick one to get started.
You can also create a list by importing a Segment Definition
--
41. Part I : Working with Recommended Audiences
--
Smartlists All Users New Users Returning Users Page(s) Viewers Converters
42. Recommended audiences (1)
Let’s see how to create one of the most common lists :
Users who visited a specific section of my site (e.g. /index.html, shirts, /cart/)
Step1 : Click on the Predefined list
Step2 : Click the Pen icon to edit
--
Lists Index
43. Recommended audiences (2)
Let’s see how to create one of the most common lists :
Users who visited a specific section of my site (e.g. /index.html, shirts, /cart/)
Step3 : Enter the page/category name*
*Note : Never enter a Full URL, but just a segment of the URI
eg. Instead of www.site.com/shoes, just enter /shoes
--
Lists Index
44. Recommended audiences (3)
Let’s see how to create one of the most common lists :
Users who visited a specific section of my site (e.g. /index.html, shirts, /cart/)
Step4 : Define Membership duration
Step5 : Name your list
Step6 : Save your List* - et voilà.
*Note : Once saved, you can’t delete a List, but you can close it (list will stop accruing users).
--
Lists Index
45. Part II : Building a custom List
--
Non-Bouncers Engaged Users High Spenders Repeat Buyers Gold Customers Past converters
46. Should you create Precise Lists ?
The more precise your lists are, the greater your ROI will be.
On the other hand, ultra-precise lists may target low volume of users.
You need to find the tipping point between Volume & ROI
* Business Intelligence
--
PRO TIP # 1
Always be using Segments in
GA Reporting to find out the
best Opportunities and set up
relevant matching conditions for
your Lists
e.g: Should you really exclude
bouncers ?
e.g: What can be considered as
a good session duration ?
PRO TIP # 2
BI* also rules on Digital.
Make sure you get a deep
understanding of your users &
converters.
Based on their intent, ads
exposure, context, timing, what is
their probability to convert ?
The more precise you are in
tailoring 1:1 conversations the
better your ROI will be.
PRO TIP # 3
Do not only focus in turning your
Users into Converters. There are
also great opportunities to re-
engage with Converters.
e.g : do you cross-sell to people
that have just purchased a
product ?
e.g : do you treat similarly loyal
vs non loyal Customers ?
PRO TIP # 4
While you can create advanced
lists with multiple conditions in
GA, it is possible to create
custom combinations in
AdWords.
This allows much better flexibility
and also to adapt on-the-fly
how you mix multiple conditions
together while preserving big
pools of user cookies.
47. The Audience Builder.
The Audience builder will allow you to easily create custom lists.
Categories for most common GA Signals are available in left part of the Builder.
Use Advanced Options
to create complex lists,
mixing various Conditions or
using Sequential Conditioning.
eg. If you want to use a Behavior
signal (like session duration), click
on the Behavior category,
then edit Matching conditions.
--
48. Using the Audience Builder (1)
Let’s see how to use these signals to create Audience Lists :
Users with Revenue above 100*
Choosing the right scope for
your matching criterias :
• User : any user that has generated
more than 100 during time frame
you’ve defined in Look-back window.
• Session : any user that has
generated more than 100 during a
single session.
• Hit : any user that has generated
more than 100 during a single
transaction.
* (Currency as defined in the GA account)
- Click on Enhanced Ecommerce** category
- Select the [Revenue] dimension & choose the scope (User, Session, Hit)
- Choose [Greater than or equal to] Matching condition
- Define its value at [100]
** If you are not yet using Enhanced commerce but the legacy version of ecommerce tracking,
then you’d need to choose the [Revenue] Dimension, by using the Conditions Option
--
Lists Index
49. Using the Audience Builder (2)
Let’s see how to use these signals to create Audience Lists :
Users spending more than 5 minutes on site
Choose the right scope for your
matching criterias :
• Per User : any user that has spent
more than 5 minutes in total during
time frame you’ve defined in Look-
back window.
• Session : any user that has spent
more than 5 minutes during a single
session.
- Click on [Behavior] category
- Choose the [Session duration]* Dimension
- Choose the [Greater than] matching condition
- Define [300] as value for the Dimension
* Note : Session duration is defined in Seconds
--
Lists Index
50. Using the Audience Builder (3)
Let’s see how to use these signals to create Audience Lists :
Users who visited the site at least 3 times
Choose the right operators for
your matching conditions :
= : Equal to
≠ : Not Equal to
> : Greater
≥ : Greater than or equal to
< : Less than
≤ : Less than or equal to
- Click on [Behavior] category
- Define [Session]* as [greater than 2]
* Note : Session are always at defined user Level during the time
frame defined by the Look-back window.
--
Lists Index
51. Using the Audience Builder (4)
Let’s see how to use these signals to create Audience Lists :
Users who added / or removed a Product from their cart
- Click on [Conditions]
- As a Dimension choose : [qty added to cart*]
- Define condition as [Greater or equal to 1]
* Note : This Dimension is only available if you use Enhanced eCommerce, otherwise you’ll need to define an Event by coding
on site when products are added/removed from the cart, and build a condition based on this event in the audience builder.
- Click on [Conditions]
- As a Dimension choose : [qty removed from cart*]
- Define condition as [Greater or equal to 1]
--
Lists Index
52. Using the Audience Builder (5)
Let’s see how to use these signals to create Audience Lists :
Users that performed Search on the Website
Regular Expressions are very
powerful to complex matching
conditions..
In that case, the expression we enter as
a matching condition means :
Everything apart from an Empty string.
May you need to re-use :
^(?!s*$).+
- Click on [Conditions] Option
- Choose [Search Term] as a dimension
- Choose ‘Matches Regex’ as a condition type
- Enter the regular expression in the field
--
Lists Index
53. Using the Audience Builder (6)
Let’s see how to use these signals to create Audience Lists :
Converters that returns to your site without Purchasing again.
Sequences are another very
powerful way to create
Matching Conditions.
You have multiple options :
Sequence start with :
- Any user interaction
- First user interaction
Also between steps, you can choose :
- is followed by
- is immediately followed by
- Click on [Sequences] Option
- In step 1, choose [Transactions] as a dimension
- Enter the matching condition : [greater or equal to 1]
- Click below on Add step button
- Choose same dimension as Step 1
- Enter the matching condition : [less than 1]
First user interaction is
linked to the very first user
Hit during the time frame
defined by the Look-back
window.
First option matches a
sequence that spans
multiple sessions or non
consecutive hits. The second
one restrict the sequence to
a single session and to
consecutive hits
--
Lists Index
54. Using the Audience Builder (7)
Let’s see how to use these signals to create Audience Lists :
Users that come from a specific campaign (e.g Upper-funnel Generic Keywords).
--
2 Options :
• Using Campaign name
If your keywords are well organized
by campaigns and that your generic
campaign names can be easily
identified.
• Using keywords
If you can’t use Campaign names, then
you could potentially create a
matching condition based on
Keywords. (in that example, we could
exclude Brand terms, while only
including traffic from Google Search)
- Click on [Traffic sources] category
- Apply your matching condition
Lists Index
55. Using the Audience Builder (8)
Let’s see how to use these signals to create Audience Lists :
Logged-In Users (when User iD is enabled)
--
If you’re using the User iD
feature from Google Analytics
in this Property, then a specific
dimension [User-ID Status] can
be used to target exclusively
users that have logged-in at
some stage during the time-
frame defined by the Look-back
window.
Note : you can create this list in
a non-User-ID view.- Click on Traffic sources category
- Choose [User-ID status] as a dimension
- Choose [Exactly matches] as matching condition
- Define Dimension value to [Assigned]
Lists Index
58. Anatomy of a GA Cookie.
_ga=1.2.3456.77777
Versioning Number
(Fixed at 1)
This field is used to figure out
the correct cookie in case
multiple cookies are setup in
different paths or domains.
This is a Timestamp*.
(When the cookie has been created)
GA Client iD
(Random number + Timestamp)
* number of seconds that have
elapsed since January 1, 1970
(midnight UTC/GMT)
--
59. Client iDs demystified
GA Client iD identify an anonymous device or browser instance. It is set in a Persistent cookie.
Randomly generated, stored in a persistent cookie and automatically sent with all hits by GA
Used by GA remarketing to tie User characteristics to Remarketing List Criterias.
The User ID feature in GA enables the analysis of groups of sessions, across devices,
using a unique, persistent, and non-personally identifiable ID string.
User iDs can’t be retargeted, but it allows user data import in GA, which can be used for Remarketing
Available in most CRM systems, a CRM iD is a unique identifier set for each identified user.
All customer information is attached to this CRM iD.
CRM iDs could be retargeted, but best use case is to enhance GA Remarketing with Data Import
USER ID
CLIENT ID
CRM ID
--
60. GA Client iD
--
User XXXXX
GA CID
23946
GA CID
19282
a Unique User may have several GA Client iDs
GA CIDs are unique per Device & per Browser
GA CID
23280
61. GA Client iD & Remarketing
--
User XXXXX
GA CID
23946
GA CID
29282
If a GA Client iD matches a List Criteria
an Advertising ID may* then be added to a Remarketing List
Session
#1
Session
#2
Session
#3
Session
#4
Session
#5
iD=28DDZLK987DHDKZ
.doubleclick.net
* As mentioned previously, if there is NO advertising cookie in Browser, then User can’t be retargeted
eg : List definition = Users with more than 2 Sessions
62. GA User iD
--
User XXXXX
GA CID
23946
GA CID
19282
The Built-in User ID feature in GA lets you associate engagement data from
multiple devices and sessions with a unique ID for Logged-In Users
GA CID
23280
User ID
ABCDE
a single User iD may be attached to various GA Client iDs
if the User is using multiple devices or browsers
63. GA User iD & Remarketing
--
User XXXXX
GA CID
23946
GA CID
29282
User iD feature allows to get User-centric reporting.
Cross-device or cross Browser Remarketing is not available
Session
#1
Session
#2
Session
#3
Session
#4
Session
#5
iD=28DDZLK987DHDKZ
.doubleclick.net
eg : List definition = Users with more than 2 Sessions
User ID
ABCDE
65. CRM Data for Remarketing
--
1 2 3 4 5
Select
Which Data
you want to
send to GA
Create
Custom
Dimensions to
store this data
Identify
Users on your
Site or App
Send
Custom data
to GA
Build
Remarketing Lists
Enhanced Remarketing
WITH DATA FROM YOUR CRM
Steps described above will require
ADDITIONAL CODE ON SITE + USE OF CUSTOM DIMENSIONS IN GA
66. Plan your Strategy
--
User XXXXX
Customer type :
GOLD
Lifetime Value :
3000
Last Purchase :
August 2014
You can create up to 20* Custom Dimensions in GA
Custom Dimension 2 Custom Dimension 3 Custom Dimension 4
*Up to 200 in Google Analytics Premium
STEP 1 : Select which CRM Data could be useful for Reporting & Remarketing
67. Prepare GA to store Data
--
STEP 2 : Create Custom Dimensions in Google Analytics
Within GA Management interface :
• Go in Admin > Property > Custom definitions
• Click on Custom Dimensions
• Click on ‘New Custom Dimension’
› Create your first custom dimension
› Make sure to choose [user] for its scope
68. Select how you’ll identify Users
--
User XXXXX
Customer type :
GOLD
STEP 3 : identify Users on your Site/App
with Built-in User iD feature, or a custom Dimension that could host your CRM IDs
CRM ID
882729
Lifetime Value :
3000
Last Purchase :
August 2014
The 2 Options are very similar but will have
implications on WHEN you can add Users
to a Remarketing List
Custom Dimension 1
Custom Dimension 2 Custom Dimension 3 Custom Dimension 4
User ID
ABCDE
or
- Option 2 -
- Option 1 -
69. Identify Users with GA
--
USER ID Built-In Feature
Easy to implement but
Non-Persistent across non-logged-in Sessions
Customize your Tracking code
for GA to identify Users
CRM iD as a Custom Dimension
Requires 1 additional step in GA but
Persistent across all Sessions
Customize your Tracking code
for GA to identify Users
<script>
(function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){
(i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o),m=s.
getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,
m)})(window,document,'script','//www.google-analytics.com/analytics.js','ga');
ga('create', UA-XXXXXXX-XX , { 'userId': 2323237 });
ga('send', 'pageview');
</script>
<script>
(function(i,s,o,g,r,a,m){i['GoogleAnalyticsObject']=r;i[r]=i[r]||function(){
(i[r].q=i[r].q||[]).push(arguments)},i[r].l=1*new Date();a=s.createElement(o),m=s.
getElementsByTagName(o)[0];a.async=1;a.src=g;m.parentNode.insertBefore(a,
m)})(window,document,'script','//www.google-analytics.com/analytics.js','ga'); ga
('create', 'UA-XXXXXX-XX', 'auto');
ga('send', 'pageview', {
'dimension1’: ‘2323237’ });
</script>
Important : For both solutions you should use a unique, persistent, and non-personally identifiable ID string.
70. Prepare GA for Data Import (1)
--
STEP 4 : Prepare GA to receive your Data
4.1 : Creating a Data Set
• Go in Admin > Property > Data Import
• Click the [New Data Set] Button
• Choose [User Data]
• Name your Data set & select the relevant GA view
71. Prepare GA for Data Import (2)
--
STEP 4 : Prepare GA to receive your Data
4.2 : Defining your Data schema
When you create a Data Set, you define a schema, which
is the structure that joins the data you upload with the
existing data in your hits. A simple schema consists of a
key dimension (the "key") and an import dimension or
metric. To import data, Google Analytics looks for key
values in hits that match key values in the uploaded data.
When a match is found, the additional dimension and
metric values associated with that key are added to the
existing hit data.
In our example, the key could be :
- Option 1 : the User iD
- Option 2 : the CRM iD (stored in a custom dimension)
The imported Dimension would be [Customer Type]
72. Prepare GA for Data Import (3)
--
STEP 4 : Prepare GA to receive your Data
4.3 : Uploading your Data
Manual Upload
You just need to upload a correctly formatted CSV file, e.g :
-----
gadimension1,gadimension2
12345,gold
45678,silver
Upload via the API
If you want to automate Data Upload, you
could use the API to regularly send User data.
You’d need to use the custom data source iD
provided when you create the schema & send
correctly formatted data.
73. Option1 - UID & Remarketing
--
User XXXXX
If you attach user Data using the Built-in User iD feature in GA,
You could retarget Users - after Data Import - as soon as they Login again.
User ID
ABCDE
DATA
IMPORT
GA CID
23946
GA CID
29282
User ID
ABCDE
Customer type
GOLD
Session
No login
Session
with Login
Session
with Login
List definition = Gold Customer
List definition = Gold Customer
74. Option 1 - Setup
--
User XXXXX
Customer type :
GOLD
Custom Dimensions act as Containers to host additional Data in GA.
You can set their values from your site and through Data import
User ID
882729
Lifetime Value :
3000
Last Purchase :
August 2014
● Step 2 : When Users log in to your site, send User iD value to GA
● Step 1 : Create custom dimensions in GA
to store your CRM User Data
● Step 3 : Attach Additional information
from your CRM using Data Import
75. Option2 - CRM iD & Remarketing
--
User XXXXX
If you attach user Data using a Custom Dimension (CRM iD) in GA,
You could retarget Users - after Data Import - as soon as they come back to the site.
CRM ID
882729
DATA
IMPORT
GA CID
23946
GA CID
29282
CRM ID
882729
Customer type
GOLD
Session with
or without Login
Session with
or without Login
List definition = Gold Customer
List definition = Gold Customer
76. Option 2 - Setup
--
User XXXXX
Customer type :
GOLD
Custom Dimensions act as Containers to host additional Data in GA.
You can set their values from your site and through Data import
CRM ID
882729
Lifetime Value :
3000
Last Purchase :
August 2014
● Step1 : Create a first Custom Dimension in GA called CRM iD
● Step2 : Create additional custom dimensions
to store your CRM User Data in GA
● Step3 : When Users log in to your site,
send CRM iD value to GA
● Step4 : Once their CRM iD is set in GA
Attach Additional information from your CRM
using Data Import
77. CRM Data for Remarketing
Let’s see how to use CRM Data to create Audience Lists :
Users with Customer Type = Gold
Users Eligibility :
For any matching users to enter
your Remarketing list, we need
a subsequent visit to your site
after that you have uploaded
Custom Data
- Click on [Conditions] category
- As a dimension, choose [Customer type]*
- As condition rule, select [exactly matches]
- As dimension value, choose [Gold]*
--
* Note : these are examples of Custom Dimensions names.
Your own custom dimensions & values may have different names
Lists Index
78. CRM Data for Remarketing
Let’s see how to use CRM Data to create Audience Lists :
Users with a Lifetime value between 1000 & 4999
Custom Dimensions & Numbers :
Custom dimensions values are stored
as Text strings. To use them as
matching criterias, we can’t use the
standard number operators. Regular
expressions are a great help here as
we can use them to match Number
ranges.
The Regular expression we’ve used
here indicated that first digit is between
1 & 4, and that 3 digits coming after
are between 0 & 9.- Click on [Conditions] category
- As a dimension, choose [Lifetime value]*
- As condition rule, select [Matches Regex]
- As dimension value, input : [1-4][0-9][0-9][0-9]
--
* Note : these are examples of Custom Dimensions names.
Your own custom dimensions may have different names & values
Lists Index
79. CRM Data for Remarketing
Let’s see how to use CRM Data to create Audience Lists :
Users that have made their last purchase in July 2014
Dates as custom dimensions:
Once Membership duration of your list
is expired, users can’t be retargeted
based on their previous activities.
On the other hand, a custom
dimension set at User level will remain
active till changed, or till GA cookie
expiration date (which is set at 2 years
after the last visit).
Note : to add again such users to a
list, they would need to visit the site
after the custom dimension value has
been imported in GA
- Click on [Conditions] category
- As a dimension, choose [Last purchase date]
- As condition rule, select [Matches Regex]
- As dimension value, input : [0-3][0-9]072014
--
Lists Index
81. Measurement Protocol
SEND STANDARD & CUSTOM DATA TO GOOGLE ANALYTICS
from any internet-connected device
This is particularly useful to report back to GA
OFFLINE EVENTS & TRANSACTIONS
but also for Email openings
--
82. MP* & GA Client iD
http://www.google-analytics.com/collect?v=1&tid=UA-XXXXXX-X
&cid=3456.77777&t=pageview&pa=purchase&ti=52ea5&tr=1100&ds=paris-store
Any Hit sent through the Measurement Protocol must contain a
GA Client iD that identify a specific User
User YYYYY
GA CID
112.888
GA CID
3456.77777
Important : a same User may have multiple GA CIDs,
you need to choose one of them and attach it to
Measurement Protocol hits if you plan to Retarget the
user based on Measurement Protocol hits data.
* Measurement Protocol
GA Client iD Hit Type Data Source
--
83. Offline Events & Remarketing
Any Hit sent through the Measurement Protocol must contain a GA Client iD
The device/browser with a same GA Client iD can be retargeted,
after a subsequent visit
User YYYYY
GA CID
3456
GA CID
1128
Session
#1
Session
#2
with Measurement Protocol
MP HIT :
OFFLINE
PURCHASE
CID:3456
List definition = Offline Purchase
--
84. MP* Hits & Remarketing Eligibility
Depending on when the on-site visit & the MP hit are happening,
Users could potentially be added into a Remarketing List
without any subsequent visit
GA CID
1234
Session
#1
Session
#2
* Measurement Protocol
List definition = Offline Purchase
GA CID
5678
MP Hit
Purchase
MP Hit
Purchase
2 hours
15 days Session
#2
List definition = Offline Purchase
25 days
Session to Hit
Hit to Session
Eligible Time Periods :
• GDN : 10d / RLSA : 6h
• Up to 30d (Look-back window)
--
85. Remarketing & Offline events
Let’s see how to use CRM Data to create Audience Lists :
Users that have purchased Shoes in the Paris physical store
- Click on [Conditions] category
- Use [Sessions] for the List scope
- Choose [Data source] as a first dimension
- Set its value to [Paris-store]*
- Add a AND operator
- Select a 2nd dimension: [Transaction]
- Select [per Session] for its scope
- Set a [Greater than or equal to] rule
- Set 1 for its value
- Add a AND operator
- Select a 3rd dimension [Prod. category]
- Define its matching rule to [Contains]
- Set its value to [Shoes]
*Important :
This setup is just an example. Dimensions & Values you’ll
use depend on how you define your Measurement
Protocol Hits.
--
Lists Index
86. MP* & Emailing
Measurement Protocol can be used where standard GA implementation can’t work
e.g : Email Openings could be tracked in GA
* Measurement Protocol
Track email openings in GA by placing an image in the content of the email,
& personalize the CID value for each of your email recipients.
<img src=”http://www.google-analytics.com/collect?v=1&tid=UA-XXXXXX-X
&cid=3456.77777&t=event&ec=emailing&ea=opening&el=Fall2015”>
GA CID
3456.77777
--
87. Remarketing & Email openings
Using Measurement protocol Hits to create Audience Lists :
Users that have opened an email but who have not visited the site afterwards
Custom Combo in AdWords:
To target this specific audience, we’ll
need to create a custom combo in
AdWords.
List A contains Email openers
List B contains Email openers that have
then visited the site
[List A] - [List B] will then exclusively
target Users that have NOT visited the
site afterwards.
--
List A
List B
LIST A
LIST B
Note : This will work if Users are in an eligible Session to Hit time frame (See this slide)
Lists Index
89. Cross-Device Remarketing
Cross Device Remarketing
could potentially be deployed, but it requires :
- Usage of Custom Dimensions
- Data Import
- A subsequent visit from another device after Data import
- That you can identify this user on this other Device or that this
user has already logged-in on this other device and that you’
ve setup his CRM iD in a custom Dimension
- That your Lists contain enough users to start serving
--
90. in-a-Nutshell ...
User YYYYY
GA CID
3456
GA CID
1128
Session
#1
Session
#2
Cart
Abandonment
DATA
IMPORT
GA CID
1128
CRM ID
ABCDE
Customer type
CART
ABANDONER
CRM ID
ABCDE
Session
No Login
Session
with Login
GA CID
9999
Session
with Login
GA CID
9999
Session
with or
without
Login• Device from where cart has been abandoned can be targeted right away
• Other devices without any login before data import would need a subsequent Logged-in visit
• Other devices where users had already login, and where GA cookie is still present could be
retargeted after data import as soon as there is a subsequent visit (with or without Login)
--
91. Cross-Device Remarketing
--
Leveraging Data import for Cross-Device Remarketing
e.g : Cart Abandoners
- Click on [Conditions] category
- Select [Users] for the List scope
- Choose [CustomerType] as a dimension
- Set a [exactly matches] rule
- Set [Cart Abandoner] for its value
Note: you may have different dimension
names & values as this is coming from
custom implementation of GA
*Important :
To Retarget Users on different devices you always need a
visit to occur on these devices after that Data Import has
been performed (See previous slides).
Lists Index
93. GA (re)marketing for Apps
Finely tune who promotions and monetization campaigns are targeted to
using GA Audiences within your Admob Management interface.
IAP Ads in-App Remarketing
Powered by
--
94. GA identifier for Apps
*UUID means Universally Unique IDentifier
cid=35009a79-1a05-49d7-b876-2b884d0f825b
We call this a
UUID*
This is how
GA for APPs
recognize
Unique Users.
The UUID
is randomly generated
for each instance of an
APPLICATION INSTALL
--
95. Apps iDs Matching
User ZZZZZ
UUID
1128
Session
#1
Session
#2
Session
#3
ADID=28DDZLK987DHDKZ
eg : List definition = Users with more than 2 Sessions
Display Ads
within Your App
Retarget Users
on the Admob Network
Audiences Lists for Apps:
GA Remarketing for Apps uses a matching between
Advertising iDs (AdiD & iDFA) with the GA UUID.
- Lists can have a duration or 540 days.
- Look-back window is also available for Apps Remarketing.
- Privacy policies must be updated, as in standard Web.
--
96. AdMob Audience Builder
The Admob Audience List Builder makes it
easy for developers to build audience lists
using familiar and powerful segmentation
tools right from the AdMob front end.
Once created, these lists can be used to
better target house ads, in-app purchases,
and monetization ads all based on the user
behavior.
--
Important :
You need to connect your Admob Account to
your Google Analytics account to leverage
GA Audiences
97. In-App Marketing
Let’s see how to create an Audience List based on App Data
Users that have spent more than 5 minutes in your App
- Click on [Conditions] category
- Select [Users] for the List scope
- Choose [Time on Screen] as a dimension
- Select [per User] for its scope
- Set a [Greater than or equal to] rule
- Set 300 for its value
Create Lists in
--
Lists Index
98. In-App Marketing
Let’s see how to create an Audience List based on App Data
Users that are equipped with version 1.5 of your App
- Click on [Conditions] category
- Select [Users] for the List scope
- Choose [App Version] as a dimension
- Set a [contains*] rule
- Set 1.5 for its value
* You could set a more restrictive rule, such
as [Matches exactly] if other app versions
(e.g 1.5.5) also contain your dimension
value
--
Create Lists in
Lists Index
99. In-App Marketing
Let’s see how to create an Audience List based on App Data
Users that have failed twice to pass first levels of your game
- Click on [Sequences] category
- Select [Users] for the List scope
- Choose [Event label] as a dimension
- Set a [matches regex] rule
- Set Level(d*)?[0-5]$ for its value
- Add a AND operator
- Choose [Event category] as a dimension
- Set a [contain] rule
- Set [Failed] for its dimension
- Choose a [is Followed by] rule
- Duplicate the above conditions
--
Create Lists in
Lists Index
101. Lists Index
--
• Users who visited a specific section of the site
• Users with revenue above 100 during a 30 days time frame
• Users spending more than 5 minutes on the site during a 30 days time frame
• Users who visited the site at least 3 times during a 30 days time frame
• Users who added a Product to their cart (or removed a product from their cart) during a 30 days time frame
• Users that performed Search on the Website
• Converters that return on the site without purchasing again
• Users coming from specific campaigns (e.g Generic Search campaigns)
• Logged-in Users
• Gold Customers (CRM Integration)
• Users with Life-time value between 1000 & 4999 (CRM Integration)
• Users with last purchase in July 2014 (CRM Integration)
• Users that have purchased Shoes in the Paris physical store
• Users that have opened an email but who have not visited the site afterwards
• Cart abandoners (CRM Integration for Cross Device Remarketing)
• Users that have spent more than 5 minutes in your App (in Apps Marketing)
• Users that are equipped with version 1.5 of your App (in Apps Marketing)
• Users that have failed twice to pass first levels of your game (in Apps Marketing)
All these lists are
Compatible with
GDN Remarketing
& with Remarketing
Lists for Search Ads
in-App
Marketing Only
102. Additional Resources
--
• Google Analytics Developer Guides
- Importing User Data in GA for Remarketing
- Data Import Developer Guide
- Custom Dimensions & Metrics Developer Guide
- Retrieving GA Client iD
• Analytics Help Center
- Google Analytics Remarketing
- Dynamic Remarketing with Google Analytics
- Data Import in Google Analytics
- Google Analytics & Admob
• External Guides & Articles
- Regular Expressions for Google Analytics
by LunaMetrics
103. Outro.
Luigi Reggiani
Head of Data Analytics, Google EMEA
@LuigiReggiani
Leverage GA to have 1:1 Conversations with your prospects & Customers.
Persuasion is an Art. Coherence is a Duty
Made in Paris - November, 2015.