Big data is getting bigger, creating more challenges and opening more opportunities for businesses. This McKinsey presentation argues that CMOs and sales leaders need to take 5 actions: harness their data, put data at the heart of the organization,
The End of Business as Usual: Rewire the Way You Work to Succeed in the Consu...
McKinsey: Understanding shifts in consumer behavior
1. June 2012
Le Web, London
Any use of this material without specific permission of McKinsey & Company is strictly prohibited
Faster than Real Time
Understanding shifts in digital consumer behavior
2. McKinsey & Company | 1
So where does all this data come from?
Exabytes of data …
1,200
1,000
400
800
200
600
0
1986 1993 2000 2007 2012
One exabyte =
4,000 x info in
U.S. Library of
Congress
95%+ now
digital, vs.
25% in 2000
3. McKinsey & Company | 2
This trend will only accelerate: BIG and BIGGER data
Internet users worldwide
2020 5 B
2010 1.9 B
Digital information in the world – videos, photos, music, texts, etc.
53 zettabytes2020
800 exabytes2010
Mobile subscribers
2020 10 B
2010 5 B
4. McKinsey & Company | 3
The rise of personal mobile phones is a key driver of this trend,
especially in emerging markets
India internet traffic by type, desktop vs. mobile, 12/08–5/12
5. McKinsey & Company | 4
A smart phone now has computing power superior to the computers
needed to send a man to the moon in 1969…
6. McKinsey & Company | 5
Data storage, exabytes
50
300
250
200
150
100
0
2007200019931986
Global installed,
optimally compressed, storage
Technology performance has increased exponentially, opening the door to
big data …
1212 million instructions per second
Overall
2010200019931986
This computational
power is equivalent
to almost 3 billion
laptops
Analog
Digital
7. McKinsey & Company | 6
… and more sophisticated analytics that help make big data relevant
for business decisions
Example analytic methods
Visualization tools
Social network mappingPredictive modeling
Neural networks
8. McKinsey & Company | 7
Big data is now better, quicker and cheaper
Limited, one-shotvs.
Massive amount
of experimentation
Laggingvs.Real-time
Laggingvs.
Sometimes ahead
of time
vs.
Separate sets of non
interoperable data
Easy mash-up
vs. Stated
Behavioral- and
intent-based
Data in nodes vs. Separate
9. McKinsey & Company | 8
5
10
15
20
25
30
35
40
45
50
55
60
65
% of buyers in category
Research online
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75
% of buyers in category
Purchase online
The second e-commerce “big bang” now touches
every product category
Still in
store
Grocery
HH Products
Digital
battleground
Furniture
Footwear
DIY Clothing
Home Décor
Health
& Beauty
Mobile
phones
Gone
to digital
Electronics
DVD/Video
Books
Video Games
Computer
HW / SW
Music
10. McKinsey & Company | 9
Online research
50%
+22%
2012
2010 41%
Percentage of respondents who used online research before buying
% Internet users, Europe
Online research is becoming a major channel for shopping related
research, especially via mobile phones
21%
+75%
2012
2010 12%
Online research via a mobile phone
11. McKinsey & Company | 10
Social networks are becoming gateways to other activities …
First place to
log in on the
computer
First place to
get access to
content
Video
discovery
tools
Percent of respondents who selected social networks
+73%
2011201020092008
19
21
17
11
+162%
2011201020092008
35
31
20
13
+130%
2011201020092008
14
109
6
12. McKinsey & Company | 11
Number of days before and after
the launch of Free mobile
-4-6 0-2 2 64
0
2,000
1,000
… which makes online communities a highly cost effective
marketing channel
Online buzz for Free mobile was much higher than any competitor…
Buzz volume; Index, as part of total normalized researches
13. McKinsey & Company | 12
The end of average – online consumer is diverse and sophisticated
Online consumer segments and attributes relative to average respondent
Size of segment, indexed
16
"Digital media junkies"
30
"Traditionalists"
13
"Digital communicators" 9
"On-the-go workers"
21
"Video digerati"
9
"Professionals"11
"Gamers"
14. McKinsey & Company | 13
2.9
2.7
2.5
2.3
2.1
Offline buyers
Online buyers
Acceptance of targeted online ads
Index
Financial situation
Declared response
In great
difficulty
In difficulty Making
ends meet
Well-off Very well-off
% of the population: 10%
The “de-averaged” online customer requires highly
targeted marketing in this data rich world
15. McKinsey & Company | 14
What if we get it wrong and get lost in zettabytes?
16. McKinsey & Company | 15
5 areas where you need to act
A wealth
of data out
there: use it
It applies
to your
organisation
Data based
decision
making
Managing
through
Big data
Real
business
impact
18. McKinsey & Company | 17
People can predict flu epidemics –
and you can watch in real time
Stomach Flu
Flu
07/11/11 13/02/1230/01/1216/01/1202/01/1219/12/1105/12/1121/11/11
0
10
20
30
40
50
60
70
80
90
100
Number of searches on
‘flu’ doubling implies
you have more than 1/3
chance of catching it in
the next 2 weeks
19. McKinsey & Company | 18
Big data permits “nowcasting,” eg. consumer product launch
and search
Index; normalized 100 at launch date
Launch
Sales
Online
searches
Intensity
Percent
Week 5Week 4Week 3Week 2Week 1One week
before
launch
0
20
40
60
80
100
120
140
20. McKinsey & Company | 19
Big data permits “nowcasting,” e.g. movie box office and social
media
Box office index vs. social mentions; 2 weeks in advance, worldwide
Tweet rate correlation
outperformed existing
market-based predictors
correlation coefficient between
a movie tweet-rate and its box
office performance
21. McKinsey & Company | 20
A whole „flora‟ of start-ups using Big data is emerging…
Crunch geolocalized data
MAP MY MOBILES
22. McKinsey & Company | 21
…with diverse approaches and objectives
Understanding
time usage
Crunch Twitter
Marketing
Investing
23. McKinsey & Company | 22
Kaggle is generating very unusual and insightful content
through large data analyses
Unusual insights
Innovative
solution for
statistical/
analytics
outsourcing
In the Eurovision
contest, Israel will
vote
disproportionately
for Belarus
If you watch a movie that
ends in a number, you
will probably think less of
it than if it had a different
title
Quality of online photos
predicted by captions:
▪ Higher-rated captioned Peru,
Cambodia, Michigan, tombs,
trails and boats
▪ Poorly-rated captioned
San Jose, mommy,
graduation and C.E.O.
24. McKinsey & Company | 23
Obama campaign and Big Data
Full team of non political tech innovators
hired to develop highly specific profiles
of potential voters
This allows strong tailoring of messages
Constant enrichment of the database
with results to tested approaches
Obama’s 2012 campaign organized as a real digital operation
25. McKinsey & Company | 24
Amazon continually runs tests and analyzes vast amounts
of data to optimize web site design – EVERY MINUTE
different versions of website
for same user in sessions 1 minute apart
25+ PhD’s
7+ parallel different versions
Statistical and empirical
Real time
26. McKinsey & Company | 25
Example of best
practice
Average for
key competitors
EBITDA 2000–10
Compound annual growth rate (%)
Granular customer insight has become
a crucial competitive differentiator
6.0
10.8
5.0
11.7
-2.2
26.6
-4.9
29.5
1.7
56.5
Media
Bank
Retailer
Insurance
27. McKinsey & Company | 26
Big Data can generate significant financial value across sectors
US health care
US retail
Europe public sector
administration
Manufacturing
Global personal
location data