Chris Hauth, Director Strategy @ affilinet international, discussed the importance of Big Data for the online marketing world in general and affiliate markting in particular. The attached presentation gives a clear statement, why this topic is key for all of us, particularly looking and evaluating the regulatory situation on EU level and their potential impact for our business.
affilinet sees itself at the forefront of this development and will together with its network of publishers and advertisers continue to push for a more intelligent, transparent and customer value generating Big Data strategy!
Unlocking the Power of ChatGPT and AI in Testing - A Real-World Look, present...
affilinet_Big Data_TactixX_20120327
1. affilinet
…thoughts on:
"how big is Big Data?"
by…
Chris R. Hauth, Director Strategy, affilinet international
03-2012
2. The questions we will address today…
What is BIG DATA and how BIG is it?
What can it mean for all of us?
Why is privacy so important and what we can do
about it?
Page 2
3. Data is getting bigger and bigger…
Worldwide Digital Data Volume
(created and stored each year)
90 2015:
7,910 EB
80
70
ExaByte
60 x 61
50
BIG DATA
40
2011:
30 1,800 EB
2005:
20 130 EB
10
-
2000 2002 2004 2006 2008 2010 2012 2014
Note: The data of 2015 estimated by IDC Page 3
Source: 2011 IDC Digital Universe Study (2011/6); MIC "Information Distribution Cencus"
4. How big is your data??? Company size is no
longer a determinant
Size DataWarehouse Revenue 2011
(in TB) (in €)
VS.
Page 4
Source: Data Unser – Wie Kundendaten die Wirtschaft revolutionieren (2012), affilinet
5. What is this thing called "Big Data" actually???
All Data that can be put into an
on-line environment… Relational
Databases no
longer sufficient…
Which surpasses our traditional
ability to effectively organize,
store and manage it…
New challenges for Unstructured Data describes any
data storage, data that is not in a database…
mining, analytics
and
interpretation… Unstructured data can be textual or
non-textual…
Page 5
Source: SearchBusinessAnalytics
6. Although Data is massively growing, it's mostly
hard to grasp and understand, it's unstructured
Size and Growth of
Data (2010)
worldwide
Share of structured
+40% Data
(from 2009)
worldwide
5%
(as of 2010)
Page 6
Source: Big Data – Powering the next Industrial Revolution, Tableau
Source: 'Future Integration Needs: Embracing Complex Data', Aberdeen Group
7. There is no reason why growth of (unstructured)
data should decrease in the future!
Social media boom The Mobile Revolution
Facebook - Active users ('04-'11) Mobile Data Connections ('04-'11)
160% CAGR 20% CAGR
E-Commerce transactions "Internet of Things"
Global eCommerce Sales ('04-'11) Connected Devices 2015 (excl. PC/Mobile)
Industrials
23% CAGR
Utilities Retail Security
Automotive Travel&Log HealthCare
Page 7
Source: Facebook Prospectus 2012, YankeeGroup Topline Data Forecast 2011, JPMorgan eCommerce Forecast
8. Selected acquisitions from eBay hint at a big
data future across all channels
Example
Magento FigCard ZVent
StubHub
Positronic
CriticalPath Appchee
Where Inc Hunch
TicketTech Milo
2001/2 2007 – 2008 2009 – 2010 2011 – Q1 2012
Stumble Upon RedLaser Zong BillSafe
Paypal WHI Solutions
GSI
Hadoop (10-20 PB) TerraData (10-20 PB)
Size of bubble = size of acquisition Product Data Shop Enhancements
Mobile Payment Social Page 8
Source: dailydealmedia, alacrastore, wikipedia, ebay
9. But Size doesn't necessarily matter: Big Data is
a complex ecosystem, navigate it wisely
- Enrichment
Profitability - Normalization
- Categorization
- New Storage
- Biz Intelligence
- Vizualization
( - ) Amount of Data
Quality
( - ) Quality
( - ) Regulation Amount of Data
Page 9
Source: affilinet
10. Customer Journey – This is what we originally
thought the journey would look like…
Idealized Customer Journey
Need recognition
1
(content sites, eMailers) Idealized customer
journey from an affiliate
network perspective
2 Information Search
(review sites)
However consumers
might complete a
3 Evaluation purchase at all stages or
linger longer
And indeed reality looks
4 Purchase
different…
Page 10
Source: affilinet
11. Biz Intelligence – Horses for Courses – Different
model-mix for different needs
Orders per Publisher type per Industry
Which publisher model
Finance works best is
dependend on the
industry segment
Telco
Retail Our advertiser KAMs or
you can steer and
influence this status
quo, but we need to
Travel provide the tools
CashBack Coupon Media Portal/ Topic Sites Page 11
Communities Source: affilinet (1 Month, Germany and affilinet only)
12. Biz Intelligence – Couponing and Cash-Back not a
danger to traditional publisher models
Pre-Order Touchpoints per Publisher Model
Topic websites beat
Topic
topic and media
Media
Media and Couponing
mainly beat similar
models
Coupon
Only Cashback mostly
Cash- beats Couponing, and
Back not other Cashbackers
CashBack Coupon Media Portal/ Price Comp Topic Sites Page 12
Communities Source: affilinet (1 Month, Germany and affilinet only)
13. Enrichment – enrich and link your Big Data to
increase the quality and make it valuable
Fit predictors
Example
Evaluations
Shares
Weather Social
Recommendations
Likes
Product
Data
Holidays
Environmental
Info Opinions
Geo-locations
Special Promotions
Connect to other players of the Value Chain
Page 13
Source: affilinet
14. Vizualization – The right visualization of Data is
the key for meaningful interpretation
Page 14
Source: LinkedIn Maps, 2012
15. Too often we see big data as a way to improve
our productivity, but that's just part of the story
What are you using Big Data for???
Increase Operational Efficiency 51% No wonder that
government get's
Inform Strategic Direction 36%
it all wrong and
Better Customer Service 27% tries to curb Data
Analytics
Identify New Products/Services 24%
Enhance Customer Experience 20%
Identify New Markets 11%
Big Data should be
Faster Go-To-Market 8%
about better
Comply with Regulations 6% services for our
customers FIRST
Other 3%
Page 15
Source: The Economist - Harnessing a game changing asset (2011)
16. Privacy and end-user trust perception is key:
BUT privacy perception differs substantially
What is privacy? (EU 27) The good news is:
The information
Visited Websites 25% our industry is
interested in is not
Taste & Opinions 27%
considered private
by the end-
customer
Photos of me 48%
Mobile Number 63%
Medical Information 74% BUT: That's just
one part of the
Financial Info 75% picture…
Page 16
Source: EU -SPECIAL EUROBAROMETER 359, 2011
17. The start and the end of big data is the complex
consumer, don't forget that…!
Your Opinion of Personalized Search? (U.S.)
Total 65% 29%
Obviously we
BAD Good other communicated
something wrong!
>50 70% 24%
Age
<29 56% 38%
High Salary 75% 19% Big Data should be
>75k about privacy and
Low Salary 45% 49% protecting
<30k individual needs
Page 17
Source: PEW Institute – PIP Serach Engine Use 2012, U.S.
18. Regulation – The New EU Data Protection
Regulation is the final point of truth for Big Data
Basic Principles of Draft from EU Commission
A General Data Protection Regulation (Verordnung,
keine Richtline): direct applicability in all EU member
states
Applicable for all "personal data", with a significant
broadening of the term "personal" (e.g. targeting)
Transparency and Controll as two guiding principles
Current timeplan
with broad implications towards role of end-user suggest final date for
implementation in 2016
Significant increase of fines (now related to all
international revenue, 2%)
However political
Applicable for all companies that do business in the necessities could
EU, also Non-EU companies
significantly speed this
No changes are currently induced regarding "E-
up!!!
Privacy" Directive (Richtlinie)
Page 18
Source: EU, affilinet
19. Regulation – What does it mean now for you???
Document, Document, Document!!!
The implications for your daily work
Non-EU based players will need to comply to the new
rules
Data protection and compliance become more
important as not only individuals can sue you
(Verbandsklagerecht)
Growing Importance and necessity of well
documented "Customer Consent" Currently the draft
means significant
More and easier accessible user rights, will mean for strains on your
you to offer customers access and controll (e.g.
operations, increase in
"Right to be forgotten", personal data portability)
bureocracy and in certain
Profiling will be harder as the definition of "personal cases the death of whole
data" is significantly broadened business models
Page 19
Source: EU, affilinet
20. What are the key take aways…
Big Data is the key decision point for our industry
and we will need to….
Put the end-customers on Prio 1 will be decisive for
the industry
Proactively work with government and interest
groups to define the customer way forward
Intelligently collaborate using the non-personal data
we have to improve customer experience
Page 20
21. … thanks for time! Enjoy the rest of your day @ Tactixx!
affilinet GmbH T. +49 89 435710 170
Chris R. Hauth M. +49 151
Jospeh-Wild-Str. 20 E.m. chauth@affili.net
81829 München www.affili.net