what's new on google - may 2013 session
DESCRIPTION
Forget about channel & device, be user centric !TRANSCRIPT
Forget about channel & device, be user centric!
Last Click is not the king
???
Day to day
I’ve bought wine online !
My buying cycle
43% Recomended
website
My buying cycle
23% Research
Smartphone, Purchased
desktop
35% Research Online
before buying offline 90%
Use multiple screens sequentially to
accomplish a task over time
The web buying cycle
Multi-device in Belgium
• Le Soir, January 2012: 21%"• Trends, March 2012: 26%"• Google & Ipsos, April 2012:
22%"• AWT, May 2012: 35%"• Mobistar October 2012:
33,6%"• GFK March 2013: 42%"
Context drives device choice
We even have data’s by hour!
Context drives device choice
Devices are personal !
00H 24H
Who’s right?
THE GEEK vs. THE FATHER
Who’s right?
THE GEEK and THE FATHER
Customer 1: Online purchase Low average basket Always connected Impulsive buyer
Customer 2:
Online purchase High average basket
Internet user Rational buyer
Internet is everywhere
Keep focus
$$$$$$$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$$$$$$$$$$$$$ $$$$$$$$$$$$$$$$$$$$$$$$
2
Don’t waste money, invest better !
Play the doctor !
Customers symptoms
They are all differents but: § Always search for a specific generic KW § Always come 3 times before converting § Always consult a specific page § Havn’t received any eMail § First touch point is always the same …
User segmentation
Customer 2 Male No Smartphone Retired
Customer 1 Female Smartphone Managing director
Prospect Male Smartphone Student
Both are customers with differents habbits
What’s the goal then?
What’s the goal then?
What’s the goal then?
What’s the goal then?
What’s the goal then?
What’s the goal then?
What’s the goal then?
« Attribution modelling helps you to be less wrong »
« Attribution modelling helps you to invest better »
« All models are as wrong as usefull »
What’s the goal then? #When was the last time you did something for the first time ?
Thank you !
?
E’s
How can AdWords help today ?
AdWords Conversion Funnel
When do People Convert ?
Actual situation
Campaign 1
Campaign 2
Campaign 3
Cross-device
That girl is a session breaker !
Purchaser who did research online & offline before purchasing
Source: Consumerbarometer
People That used a device for research before purchasing
Source: Consumerbarometer
Do you know it?
Cross-device tracking !
Not anymore !
Mass Communication is over
The right message, on the right device, @ the right moment.
What’s New ?
Win moments that matter
Key Features of Enhanced Campaigns
Smarter Ads – Sheduled Ads
Smarter Ads - Customized & extended for mobile
Smarter Ads - Located around Store
Shoe Store
Smarter Ads - Context Matters !
Delivery Map
What’s next ? Google Now !
What’s Next ? Google Now!
Conclusion
Man’s mind, once stretched by a new idea, never regains its original dimensions - Oliver Wendell Holmes -
Assigning conversion to the right channel(s)"
How to determine the right attribution model
28/05/13
Common mistake : last click attribution"
53"
Other alternatives"
54"
Linear a7ribu8on
Time decay
Shopper’s multi-channel journey"
55"Google -‐ Shopper Sciences study 2012, looking at the purchase journey for 3,000 shoppers in tech, CPG, auto and finance
Awareness Considera-on Pre-‐purchase
Define the right KPIs"
56"
Brand
Product
Message
Short list
Preference
Convic8on
Search details
Search advice
Search store
Test
Buy
Postpone
Pre / post test Exposed / control group
Ac8ons (sales) Abandon step (70%) Queries
Leads Conversa8ons
Awareness Considera-on Pre-‐purchase
Identify touch points"
57"
KPI 2012 – KPI 2011 -‐ Yearly market trend = incremental impact
Econometric modeling"Base line"Market trends"Influencing factors (awareness, promotions, distribution, ads…)""
58"
New registra8ons are influenced by various factors VW Golf econometric model"
59"
Search interest for VW Golf
New registrations VW Golf
direct indirect impact
ATL advertising for VW Golf / category
Market Seasonality
Car Scrap Bonus
Brand Loyalty VW Golf
Product lifecycle VW Golf Indirect
impact from other factors through search
organic search
paid search
Direct influencing factors"
60"
39,5%
14,5%
12,7%
15,7% 15,7%
17,7%
100%
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Basic brand loyalty (no short term effects)
Car Scrap Bonus (and seasonality)
VW Golf product lifecycle
Short term brand interest (organic search)
Advertising (TV, Press, Online / SEM, Outdoor)
Total new registrations for VW Golf
Being the category leader, over 50% of VW Golf new car registra8ons are a direct result of brand loyalty – either from a basic brand loyalty that does not move quickly or from short term changes in brand interest (indicated by an upli^ in search volume). Adver8sing accounts for 18% of registra8ons.
Total number of new registra8ons
Trommsdorff & Drüner: VW golf econometric modeling 2010
Econometric modeling"Germany : 77% online searches before car purchase vs Belgium 66%""
ROPOResearch Online Purchase Offline!
Consumer Barometer IAB Europe – TNS -‐ Google
Google benchmarks"
62"
TV ad boost = + 25%
Radio ad boost = + 33%
Influence of media on search"
63"
-‐
200.000
400.000
600.000
800.000
1.000.000
1.200.000
1.400.000
-‐
20
40
60
80
100
120
140
160
180
200
Adver8sin
g pressure (€
rate card)
Searches
Free sheets Display Mag Outdoor Dailies Radio TV Car brand search index
Day Free sheets Display Mag Outdoor Dailies Radio TV Total mediaCorrel -‐6 42 33 11 8 37 70 50 R² 0 17 11 1 1 13 49 25
Econometric modeling : example"Is search predictive of sales ?"Google Germany &marketing consultants project : predict new car registrations for VW Golf based on branded search queries. ""
uplift in queries
MONTH
1 MONTH
2 MONTH
3
2 months Hme lag
6 to 8 weeks prior to an upli^ in new car registra8ons an upli^ in search queries can be seen on Google.
Total model predic8on quality:
± 10% of quarterly new registra8ons for VW Golf
+ 10% 4.239
+ 10% + 16%
15 search queries 1 registra8on
2 months Hme lag
2 months Hme lag
2 months Hme lag
Trommsdorff & Drüner: VW golf econometric modeling 2010
Conclusions about search & media Media influence"• Direct sales (awareness, consideration)"• Searches which lead to sales (pre-purchase)"
Searches are predictive of sales, they are not responsible for all sales.""Search influence is maximal"• In pre-purchase phase to avoid last minute choice of
competitor (travel, promotional markets, easy to compare products, limited complexity, limited service)"
• In purchase phase to retarget abandoned purchase"• For product with short buy cycles or impulse buys"
65
Key take aways"• Define your KPI"• Identify touch points"• Use econometric modeling to assign a weight to
influencing factors""Short buying cycle / impulse purchase"è More weight on last click"
Long buying cycle / rational purchase"è More weight on 1st touch point (cf brand loyalty)"
66"
Optimization of Content Management For Cross-Device Web
Marketing"
I’m a (business) geek !"
Georges-Alexandre Hanin"@gahanin"
@mobilosoft"
And I don’t seem to be alone"
1,5 TV per household*"
82% of householdown a computer**"
77% of householdare connected to the web**"
42% of Bel own a smartphone***"
33% of h. "own a tablet*"*Source: http://www.datanews.be - **: spf economie - ***: GFK Retail & Technology
It means you are not communicating to a single computer"
But to a person with multiple devices"
A person, that is connected"
And moving" 79% of smartphone users have looked for
local information*"
As a result 84% have
taken action*"
*Source: http://www.ourmobileplannet.com
Agenda"
• The new challenges of Content Management"• Managing your content across devices"
1ST CONSUMERS USE SEVERAL DEVICES"
To reach the same content at different periods"
""
Office: Computer Home: iPad Transport, toilet,
queues, meetings,…
To reach different content at different periods"
""Plan for shopping
Shopping list, directions,…
Cook, read, learn, relax
To reach different content at the same time" Tv: movie & Ads
Tablet - Smartphone: more ideas
To activate content from another device at the same time"
Tablet: gameboard Smartphone: player interface
2ND CONSUMERS DO MULTI-TASKING"
*Source: http://www.ourmobileplannet.com
And they are constantly triggered"
This has an impact on your web-visits !"
*Source: http://www.ourmobileplannet.com
In conclusion: many triggers, many behaviors, many devices, many places,…"
And we haven’t talked about Social Media yet…"
Agenda"
• The new challenges of Content Management"• Managing your content across devices"
1ST WHICH CONTENT WHERE ?"
How will your users consume your content ?"
Content consumption Management Type
What to do ?
Same content at different periods
Coherently The content must be the same across all media and devices
Different content at different periods
Action-related The content must be centered around the utilization periods for each device
Different content at the same time
Complementarity
The content must be centered around the best use of each device
Activate content from another device at the same time
Simultaneity
The content must be centered around the best use of each device
2ND WHICH FORMAT ?"
Screens"+ 5.000 types of cell-phones with more than 255
different screen sizes*
Among the 11 most used tablets, there are 4 different screen sizes and 8 resolutions **
*Source: http://www.binvisions.com - **: http://www.mobilesmashingmagazine.com
Among the 30 most recent smartphones, there are 17 different screen sizes and 19 resolutions **
iPhones …"
*Source: http://www.iphoneresolution.com/
Then, you must pay attention to …"
• The device itself "• The OS & navigator"• The Internet speed
(Edge, 3G, 4G, Wifi)"• The quality of the
hardware"• The servers"• The CMS"• …"
HOW TO MANAGE YOUR CONTENT?"
1. Define your Content
Management Type "
2. Identify your target devices &
media"
3. Find the best tool to manage
your content across devices &
media "
How to identify the relevant tool ?"
Coherent Action-related
Complementarity Simultaneity
Content Consistently share data across platforms
Each content has its place (media, device)
Each content has its place (media, device)
Each content has its place (media, device)
Multiple formats
Necessary May be needed
Not really necessary
May be necessary
Performance Important Key Key Important
Publication timing
Less important Might be crucial
Key Key
A FEW WORDS ABOUT MOBILOSOFT"
We manage the virtual presence of Physical stores"
Virtual presence ? "
Promos Address & store
data
HOW DO WE DO THAT?"
We publish store data"
We create mobile websites
A mobile site for Godiva that shows what your customers really need:"ü Stores ü Promotions ü Product catalogue ü Links to social networks
Test it: http://m.godiva.be/
We make sure local promotions are well referenced"
On Foursquare, your mobile website and Google
All this is done via a single tool"
Centralize the data / content management
And synchronize across all platforms
106
« De la qualité du renseignement nait l’ac4on »
Agenda
• Scine8k
• 3 examples of the importance of the way to view DATA to get a more user centric approach in e-‐business – FROM Independent traffic sources analysis TO Smart Conversion a7ribu8on
– Mul8channel tracking on your website (and further) – User centric data to op8mize your website
1
2
3
108
Scine8k is an e-‐business consul8ng agency
109
Our Mission
Assist our client in developing profitable & long term connected business
Measuring Modeling Interpre-ng
UNDERSTAND
IMAGINE Consul-ng Innova-ng Conceiving
Realizing Op-mizing
Driving
ACT SHARE TransmiNng Guiding Coaching
data analysis
design
E-‐ marke-ng
Webanalyse
Plan de taggage
Segmenta8on
Bilans
Synergie des
canaux
Modèle économique
AMOA
Pilotage efficacité
Neuromarke8ng
Look & feel
Charte graphique
AMOA (agence créa de la marque)
Connaissance client
Repor8ngs personnalisés
Conversa8ons sociales
consul-ng Stratégie
Cross canal
Digitalisa8on magasin
CRM
Business plan
Op8misa8on con8nue
Experience client
ergonomie
Merch-‐ search-‐andising
Trigger marke8ng
Personnalisa8on dynamique
Conversa8ons et co-‐crea8on
Plan de contact personnalisé
Business
Ergonomie
WHAT WE DO
Campaign mgt
AMOA Sélec8on presta
Choix plateforme
Catalogue merchandising
112
Interna8onal mountain brand
WHO WE DO IT FOR
FROM Independent traffic sources analysis TO Smart Conversion a7ribu8on
1
Business Case
• Mission – An interna8onal mountain brand wanted to boost its business online but didn’t know which traffic sources where the best for them
• What we did – We tested different traffic sources for our client to be able to calibrate their traffic acquisi8on strategy
Similar brands:
N.B. figures presented in this business case were rebased for confidenHality reasons
1
Which traffic sources were tested ?
Traffic to website
Affilia8on: -‐ CPA -‐ CPA + CPL -‐ CPC -‐ MG
Promo codes websites / cash-‐back / cart abandonment email
Adwords Brand & generic
Emails on loyalty + prospects database
Co-‐registra8on on volume /shoppers databases
DB
1
Results
• Some sources seem to be lagging behind – Affilia8on – Co-‐registra8on
• While other sources seem to be top of class – Promo codes – Loyalty emails
• Tradi8onally, we look at how much it costs to make a sale from a traffic source (CPA)
CPA (based on last touch conversions)
34 €
5 € 11 €
46 €
3 €
-‐ € 5 € 10 € 15 € 20 € 25 € 30 € 35 € 40 € 45 € 50 €
Affilia8o
n
Prom
o code
s web
sites
Adwords
Coregistra8o
n
Emails on
loyalty
database
1
But should w
e cut the bu
dget from the
traffic sources th
at perform less
according to
this view of
the results ?
1
Actually, traffic sources don’t play the same role in the conversion process
Searched for a backpack on
google >> clicked on the
adwords ad
Monday 3 pm
Came back typing directly the url in the browser and registered to the
newsle7er
Monday 8 pm
Received an email with special promo
offer
Tuesday 11 am
Searched for a voucher on the
internet and made the transac8on
Tuesday 7 pm
Adwords Direct Email Promo code
Example :
Came back typing the exact model of
backpack and using an adword
Tuesday 7 pm
Adwords
1
How we can rebalance the results
Traffic Source Last touch Conversion
Indirect Conversion
Adwords 0 1
Direct 0 1
Email 0 1
Promo Code 1 1
Searched for a backpack on
google >> clicked on the
adwords ad
Monday 3 pm
Came back typing directly the url in the browser and registered to the
newsle7er
Monday 8 pm
Received an email with special promo
offer
Tuesday 11 am
Searched for a voucher on the
internet and made the transac8on
Tuesday 7 pm
Adwords Direct Email Promo code
Example :
Came back typing the exact model of
backpack and using an adword
Tuesday 7 pm
Adwords
1
Conversion traffic sources
• They intervene really late in the conversion process
• They make the conversion happen on your website (instead of your compe8tors)
• They take advantage of the work done by other levers before
N.B. figures presented in this business case can vary depending on your industry
Traffic Sources Conversions
345 110
250
200
850
300
180
60
280
210
0
500
1000
1500
2000
Indirect Conversions
Last touch conversions
Emails on loyalty database
Coregistra8on
Adwords
Promo codes websites
Affilia8on
1
First and middle touch traffic sources
• They intervene early in the conversion process – Clients first come to your
website thanks to them • They o^en don’t get the
commission for the conversion even though they played a role in the purchase
345 110
250
200
850
300
180
60
280
210
0
500
1000
1500
2000
Indirect Conversions
Last touch conversions
Emails on loyalty database
Coregistra8on
Adwords
Promo codes websites
Affilia8on
Traffic Sources Conversions
1
So what is the real CPA of your sources ?
• We can calculate a CPA based on indirect conversions
• Truth is somewhere in between • And if you want to be totally user centric you should look at the
traffic source as a recruitment lever and calculate the client ROI along his life
34 €
5 € 11 €
46 €
3 €
24 €
9 € 9 €
33 €
5 €
-‐ €
10 €
20 €
30 €
40 €
50 €
Affilia8o
n
Prom
o code
s web
sites
Adwords
Coregistra8o
n
Emails on
loyalty
database
CPA (based on last touch conversions) & CPA (based on indirect conversions)
1
How can you track it in Google Analy8cs ?
You can have a global view of how traffic sources are contribu-ng to the conversions And which traffic sources are complementary or not
1
How can you track it in Google Analy8cs ?
…
Or get the raw data in order to calculate your self which traffic sources are first touch / last touch for you And calculate the real ROI of your traffic sources
1
Mul8channel tracking on your website (and further) 2
Business Case Truffaut
• Mission – Truffaut (leading French gardening company) wanted to evaluate the performance of its brand new e-‐commerce + content website in terms of e-‐commerce and mul8channel ac8vity (58 stores in France)
• What we did – Set a mul8channel tracking strategy in analy8cs – Produce mul8channel reports
2
Seyng up a mul8channel tracking strategy 2
Content
E-‐commerce
Stores
Product Informa8on
Product
Defining different type of visits
Seyng up a mul8channel tracking strategy 2
Tracking them correctly in analy-cs
Product informa-on visit
E-‐commerce visit Store visit
Seyng up a mul8channel tracking strategy
• Example : – Traffic Acquisi8on strategy driven
by e-‐commerce and stores KPIs
• Color coding – E-‐commerce KPIs = orange – Stores KPIs = green
2
And producing reports with mul-channel data
User centric data to op8mize your website 3
Introduc8on to Neuromarke8ng 3
Opera8onal NeuroMarke8ng Opéra8onnel u8lise les enseignements des neurosciences, de la psychologie cogni8ve, de la psychologie évolu8onniste et de l’économie comportementale pour améliorer l’efficacité des ac8vités Marke8ng/Communica8on liées à la performance, à l’ac8on.
Le pilier fondamental du NeuroMarke8ng Opéra8onnel repose sur l’idée que la majorité de nos ac8ons, de nos comportements, ne sont pas les conséquences de processus conscients mais sont des réponses inconscientes, automa8ques, ou liées au contexte.
L´évalua8on et les recommanda8ons sont générées en analysant : • La mécanique de l’a7en8on • La mécanique de la mémorisa8on • La mécanique de la prise de décision automa8que
Introduc8on to Neuromarke8ng 3
Opera8onal NeuroMarke8ng uses Neurosciences techniques to op8mize performance driven marke8ng and communica8on ac8vi8es
The main idea of Opera8onal NeuroMarke8ng is that most of our ac8ons are not driven by a conscious mental process, but are the results of automa8c, unconscious automa8c responses from our brain
Opera8onal Neuromarke8ng 3
Example: How to design your homepage banners to maximize your CTR
-‐ Strengthen color contrasts -‐ Make people look at your CTAs (Call to Ac8on)
Opera8onal Neuromarke8ng 3
Example: How to make people read your content ?
-‐ No contrast -‐ Long lines of text -‐ Page architecture
Opera8onal Neuromarke8ng 3
Example: How to make people read your content ?
-‐ Faces aligned with the text to be read -‐ Capital le7ers -‐ Small paragraphs
Thanks
Contacts: Thibault Jaime
+33 6 72 62 18 28 [email protected]
Xavière Tallent
+33 6 73 80 47 08 [email protected]