MOVING TO CUSTOMER VALUE
WITH GOOGLE
Pawel Matkowski Global Product Lead, Analytics and Big Data MEASUREMENT AND ATTRIBUTION
USER CENTRICITY: CROSS-DEVICE WHAT ONLINE TO OFFLINE: INTRODUCTION TO OMNICHANNEL
MULTI-TOUCH ATTRIBUTION IS IT ALL
MOVING TO LIFETIME VALUE
ABOUT? MOVING TO CUSTOMER VALUE FROM CONVERSION FUNNELS TO LIFETIME VALUE
Organic Search
YESTERDAY
Display Network Organic Search
TODAY
Display Network Generic Paid Search x-device conversion
Display Network Organic Search Display Brand
TOMORROW
But when? ;-) MEASUREMENT HAS GOTTEN
MORE INTERESTING
The adoption of mobile New measurement Merging of online to offline devices technologies engagements
Single device, online measurement is no longer sufficient A T T R I B U T I O N The process, methodology and technology of assigning value to
the media that impacted the user along the customer journey
RESEARCH
RESEARCH PURCHASE AWARENESSS
PURCHASE
Display Network Generics Brand
Brand WE BREAK ATTRIBUTION INTO
THREE DIFFERENT BUCKETS
ACROSS DEVICES/ENVIRONMENTS ONLINE-TO-OFFLINE
Desktop Desktop Store Mobile App Conversion Conversion Search Search Visit
ACROSS CHANNELS
Display Search Conversion GOOGLE’S MEDIA, MEASUREMENT AND
ATTRIBUTION PLATFORMS
Attribution 360
Measures Measures Online/offline Online/offline Google Media All Media measurement marketing mix
Advertising Platforms, with Measurement Platforms, with Measurement and Attribution Ad Tracking enabled enabled C R O S S-
LET’S START D E V I C E MOST COMMON INDUSTRY SOLUTIONS
LOG-IN BASED SIGNALS BASED
(“Deterministic”) (“Probabilistic”)
First Party User Authentication Third Party User Authentication Third Party Data
User ID Cross-Device Tracking
Proprietary data, user privacy Highly scalable, user privacy Scalable solution across publishers Limited to size of signed-in user base Limited to few, large publishers Not 100% accurate, user has less control RECOMMENDATION
COMBINATION OF 1st & 3rd
PARTY LOG-IN BASED SOLUTIONS
F I R S T P A R T Y T H I R D P A R T Y
Start building your 1st party graph today Leverage 3rd Party signals to
As technology improves, value of this graph inform media bids and budgets
for measurement & targeting will grow CROSS-DEVICE The Way With Google
CROSS-DEVICE MEASUREMENT
MOVING TO USER CENTRICITY
SCALE METHODOLOGY BREADTH +1B Android Phones
ANONYMOUSLY
OBSERVE X-DEVICE CONVERSIONS ACROSS OUR
USER BASE*
+1B Monthly Actives +1B Monthly Actives *Anonymous users previously logged into Google properties on multiple devices Tackling the Cross-Device Landscape
CROSS-DEVICE AND CROSS-ENVIRONMENT
MEASUREMENT
Web across devices In-app and Web (Display Only)
Buys three pairs of User clicks on User clicks on Later orders flowers on jeans on Retailer’s Mother’s Day flowers Retailer’s desktop ad a website on his tablet mobile site ad inside gaming app EXPAND DATA Country Conversion type Date Landing page Devices used TO REMAINING USERS USING
We only surface MANY FACTORS 95% data when we have a 95% confidence
interval T A K I N G A C T I O N O N
CROSS DEVICE DATA TAKING ACTION
AUTO-BIDDING TO
CROSS-DEVICE IN ADWORDS
Google’s bidding algorithm automatically maximizes your AdWords attribution reporting now gives conversions across devices; no insights into full cross-device paths need to manually set a mobile
bid modifier U S E R I D Standard GA Tracking:
COOKIE BASED
AD SITE APP
User 1 _utma=111 &cid=111
AD SITE
User 2 _utma=222 &cid=222
AD SITE
APP
User 3 _utma=333 &cid=333 GA Tracking with User ID:
USER BASED
AD SITE APP
AD SITE
User &uid=111
AD SITE
APP GOOGLE ANALYTICS
DEVICE OVERLAP GOOGLE ANALYTICS
DEVICE PATH DATA Reebonz grew ROAS on Google Search by 55% driven by cross-device insights in Analytics
OUR APPROACH A single view of user across devices & assets on GA360: Incorporate full app and web tracking on a single property via User ID ● Key Insights: Reebonz was able to see that: ○ 36% traffic & 52% new visitors originated from mweb ○ When mWeb is part of path to purchase, CvRs improved dramatically by up to 2.8x ○ Avg. Revenue per User also increased 2.5x when customers researched on mWeb before purchasing ● Action as a result of insights ○ Included Cross-Device Conversions in DoubleClick Search Auto-bidding ○ Increased mobile bids from -60% to 0%
RESULTS Conversion Rate Average Revenue per User 55.4% increase in ROAS in SEM
20% increase in transactions in SEM
The solution and analysis methodology is now replicated across all Reebonz markets Getting Users to Log In
U S E R I D EASY
GETTING LOG INS REWARDS
GETTING LOG INS ACCESS
GETTING LOG INS LOYALTY
KEEPING THEM
Receipts from offline sales sent out to e-mails/uploaded to online profiles
Shopping and wish lists e.g., for travel, retail, etc. Personalized recommendations e.g., next trip based on the past history
Price alerts: customer reviewed product X, if price decreases by 10%, send the registered customer an alert
Product reservations for offline/online purchases
Discounts and special offers for members only
Pre-sales days for members only N O W, N E X T
What Now Next
Enable cross-device Enable User ID in bidding in Cross-Device Analytics and drive AdWords/DoubleClick loyalty Search ONLINE TO
INTRODUCTION TO OMNICHANNEL OFFLINE ONLINE TO
OFFLINE MEASUREMENT AND ATTRIBUTION
optimizing to offline sales vs. Leads to Sales online leads
the correlations between online Online to Store advertising and offline purchases Measure call/leads How to accurately
WIN WITH
CALL LEADS
Allow clients Optimize your
to easily call you call center HOW DOES IT
AD WORK TODAY?
Online conversion tracking
AD
SITE
AW CT
AD FOR SOME HOWEVER, IT WORKS DIFFERENTLY
AD
Online leads are
qualified on the phone
AD Sales close after
SITE in-person meetings
$ AW CT $ Revenue generation
AD happens in store The basic
DATA FLOW
1 2 3 4 5 User clicks on an User submits the Unique Identifier is The lead/sales is Offline data ad and arrives on GA lead (form). passed and stored in closed offline. upload. the site. your database.
AW
AD FORM
DCLK
Unique ID Unique ID 748596123 748596123 INTEGRATING OFFLINE DATA
addresses this challenge
BENEFITS
3 1 2
Understand the Optimize campaigns Improve your true return on that drive the most online targeting investment from total value, not just with offline your digital media online value intelligence Google offers several solutions to INTEGRATE OFFLINE
DATA
AdWords | DoubleClick Search | Analytics & Analytics 360 | Platform | Product Offline Conversions Offline Conversions Measurement Protocol
Import offline transactions Data Import offline transactions Import offline transactions & customer data I SunMaster Real-time call center data powering search marketing
“If your webmaster has a basic understanding of programmatic language, you can look at the transaction section of the Measurement Protocol documentation online and be good to go. This is relatively easy to set up for any business out there. We have a 99% data accuracy between Google Analytics and our internal CRM system. Via Google Analytics, transaction data gets imported back into Adwords.”
Martin Klavon, Web Developer, Sunmaster
Goals Approach Results
Correlate cross-channel marketing Recorded nuanced data using Attained a more efficient management activities with sales Custom Dimensions of marketing activity that enables greater focus on strategy Collect and integrate reliable data Employed Measurement Protocol to produce actionable insights to integrate call centre CRM Click-through rates as high as 15% information into Google Analytics and conversion rates near 50% Improve marketing strategy through Smart Lists Used Smart Lists to bid on
generic keywords on Search
How
WE’VE CHANGED
MONOCHANNEL MULTICHANNEL OMNICHANNEL
L PILLARS OF 3 OMNICHANNEL CAMPAIGNS
TARGETING MESSAGE MEASUREMENT
Compare the CTR uplift of an Creating geo-targeting campaigns Enhance messaging to drive omnichannel campaign with an around each store users to the store always on
Segment the categories Add ‘I-want-to-go’ micro-moment Add ‘I-want-to-go’ micromoment you own and assess where keywords to your capture users Keywords in your ad text to invest your marketing dollars that instant based on campaign efficiencies IDENTIFY THE USER ONLINE
Offer users real benefits of logging 1 into your site on every device - create a brand-wide loyalty strategy user login User ID (UID) assigned
Find ways to keep users logged in via 2 soft-logins, incentives and others
Implement User ID feature using a tag 1. Track authenticated User ID customers browsing online 3 management platform like Google Tag Manager - it is simple! AND OFFLINE (IN-STORE)
Create a brand-wide loyalty strategy for your customers with real benefits e.g., 1 special offers, coupons, pre-sales, discounts 2. Identify customers making
purchases at checkout in store
Find ways to identify users at checkout in store, via: loyalty cards, receipt 2 e-mailing, in-app payment systems, NFC User ID Loyalty Card purchase
Make sure that you can match the User Offline/CRM/BI Integration 3 ID assigned upon online login with in-store checkout identification Measurement Protocol
eCommerce Data In-Store Sales Data
AdWords brought $149k in online revenue and contributed to additional $479k of in-store sales. Based on the last non-direct attribution only! Case Study: Petite Bateau & Google
1 2 3
Context In-store buyers In-store buyers with loyalty with loyalty cards cards that log-in on the 153 stores in France website 36 days of store data loaded in A high % of transactions’ volumes Google Analytics are made through the loyalty card Logged-in users represent a high % of program online traffic that can be matched From 10/26/2015 to 12/1/2015 with offline transactions made with loyalty cards
Source: Lorem ipsum dolor sit amet, consectetur adipiscing elit. Duis non erat sem Proprietary + Confidential
Key Findings
44% Mobile x3 x6.4
of in-store buyers visited Is where the O2S effect O2S1 effect when AdWords ROAS when the site before making a is maximized average basket value is in-store sales are purchase high considered
1 In-store buyers who visited the site before making a purchase (the standard lookback window of this study is 7 days) N O W, N E X T
What Now Next
Get a holistic Import and optimize to performance view offline conversions in across all of your online Online to Offline AdWords via Conversion media by importing Import offline conversions to Analytics MULTI-TOUCH
ATTRIBUTION LET’S STEP UP MANY PLAYERS, BUT
LAST-CLICK WINNER TAKES ALL
Display Network Brand Generics
0% 0% 0% 0% 0% 100%
AWARENESS CONSIDERATION PURCHASE
Fewer Conversions More Conversions
Higher CPA Lower CPA
Negative ROI Positive ROI HOW SHOULD THAT
CREDIT BE DISTRIBUTED?
Display Network Brand Generics
40% 5% 5% 5% 5% 40%
AWARENESS CONSIDERATION PURCHASE TWO DIFFERENT APPROACHES
TO DISTRIBUTING CREDIT
Rules-Based Data-Driven
Attribution Attribution
Competitors with DDA:
- SmartPath
Advertisers use Mathematical models (Kenshoo) pre-defined or custom automatically apply credit for - Best Fit Attribution rules to apply credits to each interaction based on its (Adobe) - Visual IQ interactions in the path influence on the conversion - Convertro The Options at Hand
SIMPLE MODELS
Conservative growth strategy Aggressive growth strategy
Last Click Time Decay Linear Position-Based First Click
MORE ROBUST MODELS
Custom Rules-Based Data-Driven “What Model Should I Use?”:
Data-Driven Attribution Makes this Decision Simpler
Conversion
10% Rate
Difference of
EMAIL 0.2%
Conversion
9.8% Rate
VERY LITTLE impact on conversion rate “What Model Should I Use?”:
Data-Driven Attribution Makes this Decision Simpler
Conversion
10% Rate
Difference of
EMAIL 3.5%
Conversion
6.5% Rate
LARGE impact on conversion rate Allocate conversion credit based on each keyword’s incremental contribution to conversions...
MODEL: Data-Driven
12
60% 20% 20% 7.2 conversions 2.4 conversions 2.4 conversions
Credits are calculated for hundreds (or thousands) of conversion paths and will differ per path and per advertiser Google offers several solutions for
MULTI-TOUCH ATTRIBUTION
DoubleClick Analytics & PLATFORM AdWords Search Attribution 360 Campaign Manager Analytics 360*
MODELS Rules based & data driven Rules based & data driven Rules based & data driven* Data driven & predictive
Inter-Channel reporting SCOPE Search intra-channel All online channels Online & Offline media (bidding - served via DCLK)
Cross Device CROSS-DEVICE Cross Device Native Cross Device Native Cross Device Native Implementation
MEDIA PLATFORMS, W/ MEASUREMENT CAPABILITIES MEASUREMENT PLATFORMS, W/ MEDIA INTEGRATIONS Companies that have
successfully implemented
SEE RESULTS
Doubled Paid Search & tripled Display Increased sales and 25% 15% increase in bookings and performance at cost effective growth in ROI with 10% in media savings with ROAS/CPA with position-based & time position based model reverse time decay model decay model
TRAVEL, UK HOTELS, SPAIN B2B, US N O W, N E X T
What Now Next
Use Multi-Channel Test bidding to Funnels reports in Multi-Touch Data-Driven Attribution in Analytics to drive Attribution AdWords Search meaningful attribution insights
LIFETIME
VALUE MOVING TO MOVING TO CUSTOMER VALUE FROM CONVERSION FUNNELS TO LIFETIME VALUE
Organic Search
YESTERDAY
Display Network Organic Search
TODAY
Display Network Generic Paid Search x-device conversion
Display Network Organic Search Display Brand
TOMORROW
But when? ;-) How frequently?
Lifetime How many value? Why is lifetime value so hot? products? Cost of acquired customer? What ● Understand each customer’s revenue products? potential
● Determine more effectively how much can be spent on customer acquisition
● Identify which channels/platforms drive the most valuable customers LTV PRINCIPLES
PARETO PRINCIPLE
Customer Centricity $ reveals how to increase profits from your best customers, find more like 20% 80% them, and avoid over-investing in the rest
Jim Sterne, Founder, Chairman, Web Analytics Association 80% 20% LTV PRINCIPLES
PARETO PRINCIPLE IN PRACTICE
100% In these ecommerce
business examples, top 70% Top 20% of Buyers 20% of their customers generate roughly 70% of 50% Top 10% of Buyers their total revenue... Total Revenue
20% Top 1% of Buyers
0% 0% 10% 20% 30% 40% 50%
All Buyers Ozon is using Expected Lifetime Value for Smarter Bidding...
Situation For Ozon (pureplay eCommerce company) Google Remarketing was yielding lower ROAS than other solutions on the market. Ozon was considering ceasing their spend.
Solution To increase ad spend efficiency we proposed Ozon to integrate their CRM data: 1. For each of their customers estimate their Expected Lifetime Value (RFM model) 2. Link that computed customer data (non-PII) to Analytics 3. Create eLTV-grouped Audiences for export to AdWords - then customize bidding for each audience based on the expected customer lifetime value
Ozon’s Users grouped by Remarketing Audiences Expected Lifetime built off eLTV scores CRM Value ...to grow Conversion Rate by 100% and ROAS by 55%
Conversion Rate +100% Spend +150%
with ROAS +55%
After 4 months... Google Display grew and stayed at 2.5x the initial spend with higher Conversion Rate and Reach (volume) than other solutions on the market. Ozon expanded our solution beyond Display (GDN) and uses it also with Search (RLSA). N O W, N E X T
What Now Next
Enable cross-device bidding in Enable User ID in Analytics and Cross-Device AdWords/DoubleClick Search drive loyalty
Get a holistic performance view Import and optimize to offline across all of your online media Online to Offline conversions in AdWords via by importing offline conversions Conversion Import to Analytics
Use Multi-Channel Funnels Test Data-Driven Attribution Multi-Touch Attribution reports in Analytics to drive bidding in AdWords Search meaningful attribution insights
Identify your most valuable Target your high value Lifetime Value customers off your CRM customers and bid for lookalikes THANK YOU!