8447779800, Low rate Call girls in Uttam Nagar Delhi NCR
Panel: Powering Business Decision Making
1. 1
TURNING BIG DATA INTO WINNING
STRATEGIES FOR BUSINESS
DECISION MAKING
Corrine Moy, GfK
2. 2
The Big Data era has created many new opportunities
for businesses
But it isn’t necessarily a magic bullet to the challenges
businesses face in understanding their customers
Big Data is only useful to marketers if used in a smart way
So how can you turn Big Data into Smart Data? And
how can you ensure that you profit from insights it can deliver?
[ˈbig dā-tə,]
3. 3
?#!
%&
Variety
(data in many forms)
Data in different formats,
versions, from different
sources, with different
dimensionality and structure
Velocity
(data in motion)
Welcome to the fire hose.
The real-time data flow never
stops. Today's data will be
history tomorrow
Volume
(data in huge quantity)
Data in massive quantity.
"Collect first, think later".
Big Data requires sophisticated
infrastructure
Veracity
(data in doubt)
Volume doesn't automatically
increase precision. Managing
accuracy and reliability requires
significant analytical expertise
Big Data is complex…
Typical ‘Big Data’ characteristics
4. 4
…and can be misleading
So what is required?
Thorough analytics expertise to understand
limitations of Big Data, a reference frame and context
Biasedness
Big Data is often highly selective, resulting in
unknown systematic errors
Analytic limitations
Standard approaches to data processing,
modeling and analytics not suitable
Incompleteness
Data streams are "happening", not designed for
information gathering
5. 5
But we can make it simpler…
Big Data will lead to smart insights…
… if we ask the right questions
… if we do not use the mainstream approach to Big Data
with its apparent reliance on analyzing correlations
without guiding principles. Correlation does not mean
causality!
… if we intelligently integrate reference data that allows
us to understand the context and the why as well as
any bias of the data source used
… if the information we're after can be found in the
respective source, is accessible (e.g. via API) and can
be used from a legal perspective
8+5 10+3
15-2
26÷2
5+5+3
13
6. 6
Traces of
consumers
& things
Big
Data
…by turning Big Data into Smart Data
…capture and use Big Data from various sources
• Social Media data
• Point of sale data
• Mobile data
• Location data
• Internet traffic data
• Cookie data
• Client-owned data
• Transaction data
• Loyalty card & CRM data
• (Web) traffic data of
signed-in users
• Supply chain, sales and
operations planning data
…then process and analyze the data
• Process, clean, and merge it with other data for
reference to create context and thus Smart Data
• Data Scientists use advanced Big Data analytics to
ensure maximum and valid insights
Always ensuring privacy compliance
• Need to ensure 100% compliance with data
privacy laws – country specific
Big
data
Big
Data
Deep &
granular with
known error
Reference
data
Smart
Data
Combining the
best of both
worlds
7. 7
How clients use Big Data
Marketing
campaigns
Using sophisticated
analyzes on social media
messages, companies are
learning quickly what
pleases and displeases their
customers and prospects,
how messages spread in
the internet and which topics
they should keep on the
radar
Customer journey
and e-commerce
Observing and analyzing
Internet traffic and search
patterns (on smartphones,
tablets and PCs) and
viewing it context of in-store
experiences helps
companies to understand
the customer’s journey and
to make it
more profitable
Brick and mortar
retail
Retailers can track
customers through their
physical stores to enhance
the shopping experience,
predict the number of
checkout lanes needed, and
can use loyalty card data to
match customer detail to
specific product purchases
Finance
All types of financial
institutions are integrating
data from multiple sources,
and rapidly embracing new
data types to reduce fraud,
reduce revenue leakage
and ensure compliance with
laws and regulations
8. 8
We must not stop at applying analytics to Big Data. We must leverage our deep and
unique knowledge about types and structures of consumer data.
… how we can create maximum value for clients
Add context
to social media posts
Integrate consumer reasoning
with observationally
measured behavior
Use reference data
✓
9. 9
Smart Data in the real world : Purchase journeys
Therefore, marketers have new information needs….
Understand how
digital & traditional
channels interact,
what message is best,
at each step in the
buying process
Consumers are increasing-
ly using technology along
their purchase journey and
are sharing experiences in
real-time
Traditional research
can provide some
insight – but not the
complete picture
Technology allows
us to get more data on
the path to purchase –
& combine with survey
data to capture offline
world
Business question: How do I best allocate on/offline marcoms spend to maximize sales?
10. 10
Traces of
consumers
& things
Big
Data
We captured Big Data from these sources
• Behavioural Internet traffic data across websites – via
tracking software installed on the users’ devices
Big
data
Big
Data
Deep &
granular with
known error
Reference
data
Smart
Data
Combining the
best of both
worlds
And combined it with reference data
• Applied taxonomies to identify relevant data (needle in
the haystack) and add context
• Integration with rich online survey data (in-the-moment-
data) to add the why to the passively measured journey
How we did it… in the travel industry
Then processed and analyzed the data
• Applied machine learning algorithms to i.e. identify
relevance of touchpoints or the typical purchase journey
for a particular consumer group
So the client could reallocate & optimize media spend
• …by understanding the impact of online and offline
research in decision making and by identifying most
efficient channels for each stage of the Purchase Journey
11. 11
• 10,000 Users in the panel
• 2,100,000 Page Impressions at major
search engine
• 30,000,000 Navigation events overall
• 1,140,000,000 Server requests
• 1,500,000,000 MB of textual content
• Categorize 1,322 Websites as
Retailers, Aggregators,
Accommodation, Airlines,
Destinations, etc..
• Categorize 16,011 Search Keywords
as Travel Organization,
Accommodation, Generic, etc..
• Identify relevant data on which we
can now complete analysis
• 59 Offline Bookers, 148 online
Bookers, 20 Bookers @ specific
portal
• 33,570 relevant Navigation
Events (0,1%)
One month of data in our GfK
Media Efficiency Panel
Create a taxonomy for the
travel industry
Analyze purchase journeys utilizing
machine learning algorithms for a
selected client
1 2 3The haystack Organize the haystack Find the needle
Magnitude of digital behavioural tracking data
12. 12
Key take-aways
Don’t necessarily rely on ‘Big Data’ on its own – it might not tell you the
whole story
Without understanding the consumer context, the value of Big Data for
marketers is limited
Combine consumer data with ‘reference’ data for better insights
Act SMART– to unlock the value held in your data assets, creating winning
strategies that enrich consumers’ experience & maximum business opportunity