SlideShare a Scribd company logo
1 of 119
thinks…
© 2012 – All rights reserved Fluxx Ltd.
“Ask 10 technology vendors to come
 and talk to you about big data and
you will get 15 different definitions.”
                      ZD.net


                                       thinks…
              © 2012 – All rights reserved Fluxx Ltd.
“If you ask 10 experts what Big Data
    means, you’ll get 15 different
              answers”
            TechRepublic


                                      thinks…
             © 2012 – All rights reserved Fluxx Ltd.
What kind of “experts” are they?

                                    thinks…
           © 2012 – All rights reserved Fluxx Ltd.
Conclusion:
10 People = 15 Opinions
          Eh?




                             thinks…
    © 2012 – All rights reserved Fluxx Ltd.
What is ‘Big Data’ anyway?


                                 thinks…
        © 2012 – All rights reserved Fluxx Ltd.
What is ‘Big Data’
                                     anyway?

                         thinks…
© 2012 – All rights reserved Fluxx Ltd.
BIG
 The emphasis is usually on how complex or
hard it will be to extract value from the largest
                   of datasets.
                                            thinks…
                   © 2012 – All rights reserved Fluxx Ltd.
DATA
They imply there is stuff we need to know
                in there.

             But what is it?

                                        thinks…
               © 2012 – All rights reserved Fluxx Ltd.
“There are things we know we don’t know”
              Donald Rumsfeld




                                         thinks…
                © 2012 – All rights reserved Fluxx Ltd.
Our definition

The Big Data movement is the assumption that there
is useful insight to be had from datasets that
historically were too large to meaningfully analyse.
  For example:
     Detailed click by click visitor data from internet sites
     Information about individuals’ social network interactions and
     relationships

                                                   thinks…
                          © 2012 – All rights reserved Fluxx Ltd.
A glimpse into my past




                                thinks…
       © 2012 – All rights reserved Fluxx Ltd.
Time to process
   60
   50
   40
   30
   20
   10
    0
        0         10000   20000                 30000                       40000   50000   60000
                                        KB of data

                                                           thinks…
                                  © 2012 – All rights reserved Fluxx Ltd.
Time to process
   60
   50
   40
   30
   20
   10                                       10 years later, with Moore’s Law applied
    0
        0         10000   20000                 30000                       40000   50000   60000
                                        KB of data

                                                           thinks…
                                  © 2012 – All rights reserved Fluxx Ltd.
Time to process
100
  0
      0          20000       40000         60000

                         KB of data
                                                               thinks…
                                      © 2012 – All rights reserved Fluxx Ltd.
2 years




      Time to process


100
  0
      0             20000   40000        60000
                                                                              10 terabytes
          KB of data
                                                             thinks…
                                    © 2012 – All rights reserved Fluxx Ltd.
The questions tonight are…
1. Is it as hard as they make out?
2. What kind of insight could be retrieved
   from data?

   Tell us something we don’t know….

                                       thinks…
              © 2012 – All rights reserved Fluxx Ltd.
David                 Mark                                     Suranjan
Ellis, Station 10    Bjornsgaard, Fluxx                           Som, IM Group
                           [Adventures]
                Time for some proper content




                                                 thinks…
                        © 2012 – All rights reserved Fluxx Ltd.
Who’s Afraid of Big Bad Data?
                      Or How to Stop Data Keeping You Up At Night




0, 1, 1, 2, 3,   5,     8,    13,      21,          34,             55,
Who are Station10?

        Multiplatform & multichannel data, insight and optimisation experts




0, 1, 1, 2, 3,   5,   8,   13,   21,   34,              55,
The future is data, social & multiplatform



                                                                 How prepared/unprepared
                                                                 is your organisation for the
                                                                 following upcoming trends
                                                                    over the next 5 years?




       4 out of 5 CMOs anticipate a high/very high level of data complexity over the
         next 5 years, but only half felt ready to handle it.
0, 1, 1, 2, 3,   5,   8,   13,    21,          34,              55,         Source: Econsultancy, Oct 2011
To understand data, we need to love maths!




0, 1, 1, 2, 3,   5,   8,   13,   21,   34,   55,
Who’s Afraid of Big Bad Maths?

       Companies are not hiring those people whose
       skills are on the fringes. They may well be an
       honourable person but…they’ll be told, 'You just
       don’t have the mathematical skills that are
       nowadays required.’”

                                             Eric Schmidt, Chairman, Google

0, 1, 1, 2, 3,   5,   8,   13,   21,   34,                  55,
Maths never scared this lot!




0, 1, 1, 2, 3,   5,   8,   13,   21,   34,   55,
But we are good at spotting patterns




0, 1, 1, 2, 3,   5,   8,   13,   21,   34,   55,
But we are good at spotting patterns
            14


            12


            10


             8

                                                                           Snail
             6                                                             Fibonnacci


             4


             2


             0
                      1          2         3   4         5   6         7


0, 1, 1, 2, 3,   5,   8,   13,       21,           34,           55,
How to stop Big Data
                      keeping you awake at night


0, 1, 1, 2, 3,   5,    8,   13,   21,   34,   55,
“Call her Elsie!”




0, 1, 1, 2, 3,   5,   8,   13,   21,   34,   55,
Visualise the data!




0, 1, 1, 2, 3,   5,   8,   13,   21,   34,   55,
Visualise the data!




0, 1, 1, 2, 3,   5,   8,   13,   21,   34,   55,
Not all the data will necessarily be useful or accurate




0, 1, 1, 2, 3,   5,   8,   13,   21,   34,   55,
Understand how the consumer can be “irrational”

                      SUBSCRIPTIONS                                                           SUBSCRIPTIONS
  Welcome to                                                                Welcome to
  The Economist Subscription Centre                                         The Economist Subscription Centre

  Pick the type of subscription you want to buy or renew.                   Pick the type of subscription you want to buy or renew.

  Economist.com subscription – US $59.00                                    Economist.com subscription – US $59.00
  One-year subscription to Economist.com.
                                                                 16         One-year subscription to Economist.com.
                                                                                                                                           68
  Includes online access to all articles from The Economist since           Includes online access to all articles from The Economist since
  1997.                                                                     1997.

  Print subscription – US $125.00                                0          Print & web subscription – US $125.00                        32
  One-year subscription to the print edition of The Economist.              One-year subscription to the print edition of The Economist and
                                                                            online access to all articles from The Economist since 1997.
  Print & web subscription – US $125.00                        84
  One-year subscription to the print edition of The Economist and
  online access to all articles from The Economist since 1997.

0, 1, 1, 2, 3,   5,    8,        13,             21,                  34,                              55,
It’s 1969 again, baby!




0, 1, 1, 2, 3,   5,   8,   13,   21,   34,   55,
Create an insight investigations team




0, 1, 1, 2, 3,   5,   8,   13,   21,   34,   55,
Data is power




                           "We are looking for Predictive Modeling/Data Mining Scientists
                           and Analysts, at both the senior and junior level, to join our
                           department through November 2012 at our Chicago
                           Headquarters,” read the ad.

                           "We are a multi-disciplinary team of statisticians, predictive
                           modelers, data mining experts, mathematicians, software
                           developers, general analysts and organizers -all striving for a
                           single goal:
                           re-electing President Obama."


0, 1, 1, 2, 3,   5,   8,   13,              21,                    34,                       55,   Source: Econsultancy, Feb 2012
The value of analysing big data




0, 1, 1, 2, 3,   5,   8,   13,   21,   34,   55,
Summary
        The maths behind big data should not be scary
           Especially if you start to colour some of it in!
        Learn what is inaccurate or unusual in the data
        Make sure you have the right team and right tools
         for the job
        Look to understand your consumer
        Enjoy!
0, 1, 1, 2, 3,   5,   8,   13,   21,   34,    55,
THANK YOU!

                      david.ellis@station10.co.uk




0, 1, 1, 2, 3,   5,     8,      13,         21,     34,   55,
flux[x]



      BIG DATA, SMALL STRATEGY
      Strategy by experimentation



[x] fluxx.uk.com/consults           © 2012 – All rights reserved Fluxx Ltd.
Hello                                            flux[x]

       /   Mark Bjornsgaard
       /   Director Fluxx Adventures
       /   Technology incubator
       /   Invests in our own ideas



[x] fluxx.uk.com/consults              © 2012 – All rights reserved Fluxx Ltd.
The next 30 mins…                                                    flux[x]

       / Part 1 – A brilliant theory (10mins)
            What we mean by Big Data?
            How does it really work?
            In the real world
       / Part 2 – The real story (10mins)
            The 7 things they don’t tell you about using big data
       / Part 3 – Strategy by experimentation (10mins)

[x] fluxx.uk.com/consults                                  © 2012 – All rights reserved Fluxx Ltd.
flux[x]



       WHAT WE MEAN BY BIG DATA?




[x] fluxx.uk.com/consults   © 2012 – All rights reserved Fluxx Ltd.
flux[x]




         Big data
[x] fluxx.uk.com/consults   © 2012 – All rights reserved Fluxx Ltd.
What is the data in big data?                                                               flux[x]
                                                 Sales data
                                                 (epos and
                                                   online)
                                                                Web
                                  Sensor data
                                                              analytics




                                                   Big
                            Documents                               Advertising
                             and email            Data               analytics
                                                  Sets


                                  Unstructured                  CRM
                                   social data                databases
                                                 Machine
                                                  data –
                                                 website
                                                   logs


[x] fluxx.uk.com/consults                                                         © 2012 – All rights reserved Fluxx Ltd.
How does data become big?                                      flux[x]

                                  Volume




                       Velocity            Variety
[x] fluxx.uk.com/consults                            © 2012 – All rights reserved Fluxx Ltd.
A definition…                                                               flux[x]




   “Data that’s an order of   “Data that exceeds the          “A set of technology
   magnitude bigger than we   processing capacity of          advances that have made
   are accustomed to”         conventional database           capturing and analysing
                              systems. The data is too        data at high scale and
                              big, moves too fast, or         speed vastly more
                              doesn’t fit the structures of   efficient”
                              our database architectures”
[x] fluxx.uk.com/consults                                         © 2012 – All rights reserved Fluxx Ltd.
Like its 1996 all over again…                 flux[x]




[x] fluxx.uk.com/consults           © 2012 – All rights reserved Fluxx Ltd.
flux[x]



       HOW DOES IT REALLY WORK?




[x] fluxx.uk.com/consults   © 2012 – All rights reserved Fluxx Ltd.
How does it work?                                                                flux[x]
                      Big data is…..“ A set of technology advances
                      that have made capturing and analysing data
                      at high scale and speed vastly more efficient”




      Data scientists                         R                            Hadoop
[x] fluxx.uk.com/consults                                              © 2012 – All rights reserved Fluxx Ltd.
How does it work?                                    flux[x]
                            “big data”




                               Your data




[x] fluxx.uk.com/consults                  © 2012 – All rights reserved Fluxx Ltd.
How does it work?                                                                             flux[x]
                                        “big
                                        data”


                            6
                            5
                            4
                            3
                            2
                            1
                            0                Your data
                                Category 1   Category 2   Category 3   Category 4




[x] fluxx.uk.com/consults                                                           © 2012 – All rights reserved Fluxx Ltd.
flux[x]



      IN THE REAL WORLD




[x] fluxx.uk.com/consults   © 2012 – All rights reserved Fluxx Ltd.
In…retail                                           flux[x]

    Then – have a sale      Now – target individuals




[x] fluxx.uk.com/consults                 © 2012 – All rights reserved Fluxx Ltd.
In…online marketing                                   flux[x]
                            Now – triangulate multiple
    Then - pipeline         data sets to identify targets




[x] fluxx.uk.com/consults                   © 2012 – All rights reserved Fluxx Ltd.
In…IT                                                flux[x]
    Then – data locked in   Now – rich, real time
    unusable formats        analytics




[x] fluxx.uk.com/consults                  © 2012 – All rights reserved Fluxx Ltd.
In…risk management                                    flux[x]
                              Now – telematics and
    Then – paper validation   unstructured data




[x] fluxx.uk.com/consults                   © 2012 – All rights reserved Fluxx Ltd.
In…law enforcement                                 flux[x]
                            Now – crime hotspot
    Then – gut instinct     identification




[x] fluxx.uk.com/consults                © 2012 – All rights reserved Fluxx Ltd.
In…medical research                                  flux[x]
                              Now – semantic and NLP
    Then – keyword searches   “understanding”




[x] fluxx.uk.com/consults                  © 2012 – All rights reserved Fluxx Ltd.
In…sports                                               flux[x]

    Then – subjective “truth”   Now – objective insight




[x] fluxx.uk.com/consults                     © 2012 – All rights reserved Fluxx Ltd.
flux[x]



       PART 2
       IN THE REAL WORLD



[x] fluxx.uk.com/consults   © 2012 – All rights reserved Fluxx Ltd.
flux[x]




[x] fluxx.uk.com/consults   © 2012 – All rights reserved Fluxx Ltd.
flux[x]




[x] fluxx.uk.com/consults   © 2012 – All rights reserved Fluxx Ltd.
90% of colleagues don’t really care                 flux[x]




[x] fluxx.uk.com/consults                 © 2012 – All rights reserved Fluxx Ltd.
A Vanguard doesn’t really exist                 flux[x]




[x] fluxx.uk.com/consults             © 2012 – All rights reserved Fluxx Ltd.
Use cases are still limited                                                           flux[x]

                                            Digital
                                           marketing



                            x   Data
                            retention, d
                                 ata
                             archiving
                                              Big
                                                              Data
                                                          exploration
                                                              and
                                                           discovery
                                                                        x
                                             Data



                                                                        x
                            Machine-                         Fraud
                            generated                      detection
                               data                           and



                            xanalytics        Social
                                             network
                                               and
                                           relationship
                                             analysis
                                                       x
                                                          prevention




[x] fluxx.uk.com/consults                                                   © 2012 – All rights reserved Fluxx Ltd.
Hires are scarce…                     flux[x]




[x] fluxx.uk.com/consults   © 2012 – All rights reserved Fluxx Ltd.
R and Hadoop aren’t for all techies…                 flux[x]




[x] fluxx.uk.com/consults                  © 2012 – All rights reserved Fluxx Ltd.
flux[x]
    Big data only shows us what’s not obvious…




[x] fluxx.uk.com/consults               © 2012 – All rights reserved Fluxx Ltd.
So…using big data actually means….                                 flux[x]

       /   Colleagues who are biologically pre-programmed not to care
       /   No best practice to rely on
       /   Very specific use cases within business
       /   Hiring aching expensive people
       /   To use software anyone with an IQ <150 shouldn’t go near
       /   All in the cause of…..looking for non conclusive, obscure
           patterns, which may or may not lead to any business value




[x] fluxx.uk.com/consults                                © 2012 – All rights reserved Fluxx Ltd.
flux[x]




[x] fluxx.uk.com/consults   © 2012 – All rights reserved Fluxx Ltd.
flux[x]



      PART 3
      STRATEGY BY EXPERIMENTATION



[x] fluxx.uk.com/consults   © 2012 – All rights reserved Fluxx Ltd.
Strategy by experimentation                 flux[x]




[x] fluxx.uk.com/consults         © 2012 – All rights reserved Fluxx Ltd.
1. Set aside some time (and budget)                 flux[x]




[x] fluxx.uk.com/consults                 © 2012 – All rights reserved Fluxx Ltd.
2. Find a partner you trust                 flux[x]




[x] fluxx.uk.com/consults         © 2012 – All rights reserved Fluxx Ltd.
3. Get everyone involved                 flux[x]




[x] fluxx.uk.com/consults      © 2012 – All rights reserved Fluxx Ltd.
4. Identify a business issue you have…                                                   flux[x]

                                                    What kind of
                      Why is the ROI              people make the
                       on my CPC                      largest
                        account                     purchases?
                       declining?


                                                                There’s always a
                                    How can I drive              sales spike on
                                         sales of                  Thursday
                                   recommendation               evenings – why?
                                   s in the check out
                                         process?




[x] fluxx.uk.com/consults                                                      © 2012 – All rights reserved Fluxx Ltd.
5. Identify data sets which might hold                                                       flux[x]
    clues                     Adwords data
                          Social data

                                                                                   Web analytics
                                                                                      data


                                        There’s always a sales
              Store footfall              spike on Thursday
                                           evenings – why?
                                                                                      Brand tracking




                      Transactional
                          data                                   Logistics files


[x] fluxx.uk.com/consults                                                          © 2012 – All rights reserved Fluxx Ltd.
6. Play detective                                     flux[x]




           Patterns         Relationships          Correlations


[x] fluxx.uk.com/consults                   © 2012 – All rights reserved Fluxx Ltd.
7. Rinse and repeat                   flux[x]




[x] fluxx.uk.com/consults   © 2012 – All rights reserved Fluxx Ltd.
8. Re-humanise the story…                 flux[x]




[x] fluxx.uk.com/consults       © 2012 – All rights reserved Fluxx Ltd.
9. Shout about it                     flux[x]




[x] fluxx.uk.com/consults   © 2012 – All rights reserved Fluxx Ltd.
10. Enjoy the journey                 flux[x]




[x] fluxx.uk.com/consults   © 2012 – All rights reserved Fluxx Ltd.
To summarise                                             flux[x]

       / Define big data for yourself – see how it could
         apply to you
       / Accept the challenges
       / Start small, grow and publicise success



[x] fluxx.uk.com/consults                      © 2012 – All rights reserved Fluxx Ltd.
flux[x]



      THANK YOU!




[x] fluxx.uk.com/consults   © 2012 – All rights reserved Fluxx Ltd.
Fluxx-Don't be a (big) data hater event-181012-main presentation
How not to eat
an elephant
Suranjan Som | Head of BI Practice
Fluxx-Don't be a (big) data hater event-181012-main presentation
“As consumer
technology innovation
and a culture of social
decision-making
changes the way that
our customers make
decisions about their
purchases, it is
becoming
increasingly difficult
to differentiate our
products and services
from those of our
competition”
Customer
Centric
Big
Data
Customer intelligence has
historically relied on
assumptions based on
transactional behavior
analysis
Fluxx-Don't be a (big) data hater event-181012-main presentation
Fluxx-Don't be a (big) data hater event-181012-main presentation
Fluxx-Don't be a (big) data hater event-181012-main presentation
“Big Data presents us
with some very
exciting short and
medium term wins
around customer
insight. But the most
strategic opportunity
for us is in the
integration of Big
Data to enrich the
intelligence we hold
around our single
view of customer”
Fluxx-Don't be a (big) data hater event-181012-main presentation
Business Transformation & Enablement

Organisational Governance




                                                                                                                Compliance & Regulatory
                                                                                       Information Governance
                                    Information Delivery




                                                                   Solution Delivery
                                    Information Services


                                     Information Stores


                                   Information Integration


                                       Infrastructure
Fluxx-Don't be a (big) data hater event-181012-main presentation
Fluxx-Don't be a (big) data hater event-181012-main presentation
Fluxx-Don't be a (big) data hater event-181012-main presentation
Outsource Risk
                   Specialist Business Users

Blended Resource
                        Data Scientists
Opportunity


                   Data Services & Operations



                     Technology Specialists
Data Production             Data Storage & Management               Data Exploitation




1. Data Acquisition         2. Data Transformation       3. Data Enrichment     4. Data Dissemination




                                           5. Data Disposal




                                          6. Data Governance




                                     7. Data Lifecycle Management
Validity
Analytics                     Multi-platform Delivery          Document Management




                   Business Abstraction Layer


        Big Data                             Relational Data




    Unstructured Data                       Structured Data         Documents
Fluxx-Don't be a (big) data hater event-181012-main presentation
Fluxx-Don't be a (big) data hater event-181012-main presentation
“Big data” is based on a single fear:
   There might be something powerful
        and useful in all that data!

                                         thinks…
                © 2012 – All rights reserved Fluxx Ltd.
“There are things we know we don’t know”
              Donald Rumsfeld




                                         thinks…
                © 2012 – All rights reserved Fluxx Ltd.
So the holding position has been
hang on to it, otherwise we’ll regret it later


                                          thinks…
                 © 2012 – All rights reserved Fluxx Ltd.
thinks…
© 2012 – All rights reserved Fluxx Ltd.
thinks…
© 2012 – All rights reserved Fluxx Ltd.
So the holding position is


A total boon for the storage industry!


        But equally,… it’s hard to argue with!




                                          thinks…
                 © 2012 – All rights reserved Fluxx Ltd.
But our (real) experts have told us:
  1. It doesn’t have to be big to be useful
  2. Big doesn’t always mean complex
  3. The insight is worth it



                                          thinks…
                 © 2012 – All rights reserved Fluxx Ltd.
In other words
Size isn’t important, it’s what
  you do with it that counts




                                           thinks…
                  © 2012 – All rights reserved Fluxx Ltd.
In other words
Size isn’t important, it’s what
  you do with it that counts




                                           thinks…
                  © 2012 – All rights reserved Fluxx Ltd.
In other words
Size isn’t important, it’s what
  you do with it that counts




                                           thinks…
                  © 2012 – All rights reserved Fluxx Ltd.
In other words
                                          Size isn’t
                                    important, it’s what
                                     you do with it that
                                           counts




                         thinks…
© 2012 – All rights reserved Fluxx Ltd.
In other words
                            Size important isn’t .
                           Matters only what you do
                                    with it.




                         thinks…
© 2012 – All rights reserved Fluxx Ltd.
thinks…
© 2012 – All rights reserved Fluxx Ltd.

More Related Content

Similar to Fluxx-Don't be a (big) data hater event-181012-main presentation

How to develop a data scientist – What business has requested v02
How to develop a data scientist – What business has requested v02How to develop a data scientist – What business has requested v02
How to develop a data scientist – What business has requested v02Data Science London
 
Smarter Analytics: Big Data and Predictive Governance
Smarter Analytics: Big Data and Predictive GovernanceSmarter Analytics: Big Data and Predictive Governance
Smarter Analytics: Big Data and Predictive GovernanceIBM Danmark
 
Social Analytics In The Enterprise
Social Analytics In The EnterpriseSocial Analytics In The Enterprise
Social Analytics In The EnterpriseAlan Lepofsky
 
Telco Big Data Workshop Sample
Telco Big Data Workshop SampleTelco Big Data Workshop Sample
Telco Big Data Workshop SampleAlan Quayle
 
Steven adler ibm big data predictions
Steven adler   ibm big data predictionsSteven adler   ibm big data predictions
Steven adler ibm big data predictionsGlobalForum
 
Developing Data Products
Developing Data ProductsDeveloping Data Products
Developing Data ProductsPeter Skomoroch
 
New ways of working & knowledge sharing - Dirk W. Bijl
New ways of working & knowledge sharing - Dirk W. BijlNew ways of working & knowledge sharing - Dirk W. Bijl
New ways of working & knowledge sharing - Dirk W. BijlVlerick Business School
 
Big Data and Fast Data – Big and Fast Combined, is it Possible?
Big Data and Fast Data – Big and Fast Combined, is it Possible?Big Data and Fast Data – Big and Fast Combined, is it Possible?
Big Data and Fast Data – Big and Fast Combined, is it Possible?Guido Schmutz
 
Ibm watson - how it works, and what it means for society beyond winning jeo...
Ibm   watson - how it works, and what it means for society beyond winning jeo...Ibm   watson - how it works, and what it means for society beyond winning jeo...
Ibm watson - how it works, and what it means for society beyond winning jeo...Rick Bouter
 
SF Data Science: Developing Data Products
SF Data Science: Developing Data ProductsSF Data Science: Developing Data Products
SF Data Science: Developing Data ProductsPeter Skomoroch
 
Big data in travel domain
Big data in travel domainBig data in travel domain
Big data in travel domainPromptCloud
 
Smarter planet and mega trends presentation 2012
Smarter planet and mega trends presentation 2012Smarter planet and mega trends presentation 2012
Smarter planet and mega trends presentation 2012Joergen Floes
 
Interesting Thing about Internet of Things
Interesting Thing about Internet of ThingsInteresting Thing about Internet of Things
Interesting Thing about Internet of ThingsDr. Mazlan Abbas
 
Agile Mumbai 2022 - Abhishek Mishra | How to fail in your AI Endeavors
Agile Mumbai 2022 - Abhishek Mishra | How to fail in your AI EndeavorsAgile Mumbai 2022 - Abhishek Mishra | How to fail in your AI Endeavors
Agile Mumbai 2022 - Abhishek Mishra | How to fail in your AI EndeavorsAgileNetwork
 

Similar to Fluxx-Don't be a (big) data hater event-181012-main presentation (20)

Moving Big Data to Big Value
Moving Big Data to Big ValueMoving Big Data to Big Value
Moving Big Data to Big Value
 
How to develop a data scientist – What business has requested v02
How to develop a data scientist – What business has requested v02How to develop a data scientist – What business has requested v02
How to develop a data scientist – What business has requested v02
 
Smarter Analytics: Big Data and Predictive Governance
Smarter Analytics: Big Data and Predictive GovernanceSmarter Analytics: Big Data and Predictive Governance
Smarter Analytics: Big Data and Predictive Governance
 
Social Analytics In The Enterprise
Social Analytics In The EnterpriseSocial Analytics In The Enterprise
Social Analytics In The Enterprise
 
Big data by_mcal
Big data by_mcalBig data by_mcal
Big data by_mcal
 
Telco Big Data Workshop Sample
Telco Big Data Workshop SampleTelco Big Data Workshop Sample
Telco Big Data Workshop Sample
 
Steven adler ibm big data predictions
Steven adler   ibm big data predictionsSteven adler   ibm big data predictions
Steven adler ibm big data predictions
 
Developing Data Products
Developing Data ProductsDeveloping Data Products
Developing Data Products
 
New ways of working & knowledge sharing - Dirk W. Bijl
New ways of working & knowledge sharing - Dirk W. BijlNew ways of working & knowledge sharing - Dirk W. Bijl
New ways of working & knowledge sharing - Dirk W. Bijl
 
DAMA Presentation
DAMA PresentationDAMA Presentation
DAMA Presentation
 
Big Data and Fast Data – Big and Fast Combined, is it Possible?
Big Data and Fast Data – Big and Fast Combined, is it Possible?Big Data and Fast Data – Big and Fast Combined, is it Possible?
Big Data and Fast Data – Big and Fast Combined, is it Possible?
 
IBM Watson-How it works
IBM Watson-How it worksIBM Watson-How it works
IBM Watson-How it works
 
Ibm watson - how it works, and what it means for society beyond winning jeo...
Ibm   watson - how it works, and what it means for society beyond winning jeo...Ibm   watson - how it works, and what it means for society beyond winning jeo...
Ibm watson - how it works, and what it means for society beyond winning jeo...
 
Watson how it works?
Watson how it works?Watson how it works?
Watson how it works?
 
Business Impact of Bad Data
Business Impact of Bad DataBusiness Impact of Bad Data
Business Impact of Bad Data
 
SF Data Science: Developing Data Products
SF Data Science: Developing Data ProductsSF Data Science: Developing Data Products
SF Data Science: Developing Data Products
 
Big data in travel domain
Big data in travel domainBig data in travel domain
Big data in travel domain
 
Smarter planet and mega trends presentation 2012
Smarter planet and mega trends presentation 2012Smarter planet and mega trends presentation 2012
Smarter planet and mega trends presentation 2012
 
Interesting Thing about Internet of Things
Interesting Thing about Internet of ThingsInteresting Thing about Internet of Things
Interesting Thing about Internet of Things
 
Agile Mumbai 2022 - Abhishek Mishra | How to fail in your AI Endeavors
Agile Mumbai 2022 - Abhishek Mishra | How to fail in your AI EndeavorsAgile Mumbai 2022 - Abhishek Mishra | How to fail in your AI Endeavors
Agile Mumbai 2022 - Abhishek Mishra | How to fail in your AI Endeavors
 

More from Fluxx

The Experiments in Design results from Dots 2016
The Experiments in Design results from Dots 2016The Experiments in Design results from Dots 2016
The Experiments in Design results from Dots 2016Fluxx
 
Fohs hmv presentation_v4_20130322
Fohs hmv presentation_v4_20130322Fohs hmv presentation_v4_20130322
Fohs hmv presentation_v4_20130322Fluxx
 
Fluxx Advent Calendar 2012
Fluxx Advent Calendar 2012Fluxx Advent Calendar 2012
Fluxx Advent Calendar 2012Fluxx
 
Fluxx Luxx : Talking luxury at Fluxx Studios
Fluxx Luxx : Talking luxury at Fluxx StudiosFluxx Luxx : Talking luxury at Fluxx Studios
Fluxx Luxx : Talking luxury at Fluxx StudiosFluxx
 
Fluxx-The Tested User event-011112-main presentation
Fluxx-The Tested User event-011112-main presentationFluxx-The Tested User event-011112-main presentation
Fluxx-The Tested User event-011112-main presentationFluxx
 
Fluxx-Don't be a (big) data hater event-181012-facts, stats, quotes and quest...
Fluxx-Don't be a (big) data hater event-181012-facts, stats, quotes and quest...Fluxx-Don't be a (big) data hater event-181012-facts, stats, quotes and quest...
Fluxx-Don't be a (big) data hater event-181012-facts, stats, quotes and quest...Fluxx
 

More from Fluxx (6)

The Experiments in Design results from Dots 2016
The Experiments in Design results from Dots 2016The Experiments in Design results from Dots 2016
The Experiments in Design results from Dots 2016
 
Fohs hmv presentation_v4_20130322
Fohs hmv presentation_v4_20130322Fohs hmv presentation_v4_20130322
Fohs hmv presentation_v4_20130322
 
Fluxx Advent Calendar 2012
Fluxx Advent Calendar 2012Fluxx Advent Calendar 2012
Fluxx Advent Calendar 2012
 
Fluxx Luxx : Talking luxury at Fluxx Studios
Fluxx Luxx : Talking luxury at Fluxx StudiosFluxx Luxx : Talking luxury at Fluxx Studios
Fluxx Luxx : Talking luxury at Fluxx Studios
 
Fluxx-The Tested User event-011112-main presentation
Fluxx-The Tested User event-011112-main presentationFluxx-The Tested User event-011112-main presentation
Fluxx-The Tested User event-011112-main presentation
 
Fluxx-Don't be a (big) data hater event-181012-facts, stats, quotes and quest...
Fluxx-Don't be a (big) data hater event-181012-facts, stats, quotes and quest...Fluxx-Don't be a (big) data hater event-181012-facts, stats, quotes and quest...
Fluxx-Don't be a (big) data hater event-181012-facts, stats, quotes and quest...
 

Fluxx-Don't be a (big) data hater event-181012-main presentation

  • 1. thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 2. “Ask 10 technology vendors to come and talk to you about big data and you will get 15 different definitions.” ZD.net thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 3. “If you ask 10 experts what Big Data means, you’ll get 15 different answers” TechRepublic thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 4. What kind of “experts” are they? thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 5. Conclusion: 10 People = 15 Opinions Eh? thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 6. What is ‘Big Data’ anyway? thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 7. What is ‘Big Data’ anyway? thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 8. BIG The emphasis is usually on how complex or hard it will be to extract value from the largest of datasets. thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 9. DATA They imply there is stuff we need to know in there. But what is it? thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 10. “There are things we know we don’t know” Donald Rumsfeld thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 11. Our definition The Big Data movement is the assumption that there is useful insight to be had from datasets that historically were too large to meaningfully analyse. For example: Detailed click by click visitor data from internet sites Information about individuals’ social network interactions and relationships thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 12. A glimpse into my past thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 13. Time to process 60 50 40 30 20 10 0 0 10000 20000 30000 40000 50000 60000 KB of data thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 14. Time to process 60 50 40 30 20 10 10 years later, with Moore’s Law applied 0 0 10000 20000 30000 40000 50000 60000 KB of data thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 15. Time to process 100 0 0 20000 40000 60000 KB of data thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 16. 2 years Time to process 100 0 0 20000 40000 60000 10 terabytes KB of data thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 17. The questions tonight are… 1. Is it as hard as they make out? 2. What kind of insight could be retrieved from data? Tell us something we don’t know…. thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 18. David Mark Suranjan Ellis, Station 10 Bjornsgaard, Fluxx Som, IM Group [Adventures] Time for some proper content thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 19. Who’s Afraid of Big Bad Data? Or How to Stop Data Keeping You Up At Night 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 20. Who are Station10?  Multiplatform & multichannel data, insight and optimisation experts 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 21. The future is data, social & multiplatform How prepared/unprepared is your organisation for the following upcoming trends over the next 5 years?  4 out of 5 CMOs anticipate a high/very high level of data complexity over the next 5 years, but only half felt ready to handle it. 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, Source: Econsultancy, Oct 2011
  • 22. To understand data, we need to love maths! 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 23. Who’s Afraid of Big Bad Maths? Companies are not hiring those people whose skills are on the fringes. They may well be an honourable person but…they’ll be told, 'You just don’t have the mathematical skills that are nowadays required.’” Eric Schmidt, Chairman, Google 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 24. Maths never scared this lot! 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 25. But we are good at spotting patterns 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 26. But we are good at spotting patterns 14 12 10 8 Snail 6 Fibonnacci 4 2 0 1 2 3 4 5 6 7 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 27. How to stop Big Data keeping you awake at night 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 28. “Call her Elsie!” 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 29. Visualise the data! 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 30. Visualise the data! 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 31. Not all the data will necessarily be useful or accurate 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 32. Understand how the consumer can be “irrational” SUBSCRIPTIONS SUBSCRIPTIONS Welcome to Welcome to The Economist Subscription Centre The Economist Subscription Centre Pick the type of subscription you want to buy or renew. Pick the type of subscription you want to buy or renew. Economist.com subscription – US $59.00 Economist.com subscription – US $59.00 One-year subscription to Economist.com. 16 One-year subscription to Economist.com. 68 Includes online access to all articles from The Economist since Includes online access to all articles from The Economist since 1997. 1997. Print subscription – US $125.00 0 Print & web subscription – US $125.00 32 One-year subscription to the print edition of The Economist. One-year subscription to the print edition of The Economist and online access to all articles from The Economist since 1997. Print & web subscription – US $125.00 84 One-year subscription to the print edition of The Economist and online access to all articles from The Economist since 1997. 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 33. It’s 1969 again, baby! 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 34. Create an insight investigations team 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 35. Data is power "We are looking for Predictive Modeling/Data Mining Scientists and Analysts, at both the senior and junior level, to join our department through November 2012 at our Chicago Headquarters,” read the ad. "We are a multi-disciplinary team of statisticians, predictive modelers, data mining experts, mathematicians, software developers, general analysts and organizers -all striving for a single goal: re-electing President Obama." 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55, Source: Econsultancy, Feb 2012
  • 36. The value of analysing big data 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 37. Summary  The maths behind big data should not be scary  Especially if you start to colour some of it in!  Learn what is inaccurate or unusual in the data  Make sure you have the right team and right tools for the job  Look to understand your consumer  Enjoy! 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 38. THANK YOU! david.ellis@station10.co.uk 0, 1, 1, 2, 3, 5, 8, 13, 21, 34, 55,
  • 39. flux[x] BIG DATA, SMALL STRATEGY Strategy by experimentation [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 40. Hello flux[x] / Mark Bjornsgaard / Director Fluxx Adventures / Technology incubator / Invests in our own ideas [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 41. The next 30 mins… flux[x] / Part 1 – A brilliant theory (10mins)  What we mean by Big Data?  How does it really work?  In the real world / Part 2 – The real story (10mins)  The 7 things they don’t tell you about using big data / Part 3 – Strategy by experimentation (10mins) [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 42. flux[x] WHAT WE MEAN BY BIG DATA? [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 43. flux[x] Big data [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 44. What is the data in big data? flux[x] Sales data (epos and online) Web Sensor data analytics Big Documents Advertising and email Data analytics Sets Unstructured CRM social data databases Machine data – website logs [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 45. How does data become big? flux[x] Volume Velocity Variety [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 46. A definition… flux[x] “Data that’s an order of “Data that exceeds the “A set of technology magnitude bigger than we processing capacity of advances that have made are accustomed to” conventional database capturing and analysing systems. The data is too data at high scale and big, moves too fast, or speed vastly more doesn’t fit the structures of efficient” our database architectures” [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 47. Like its 1996 all over again… flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 48. flux[x] HOW DOES IT REALLY WORK? [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 49. How does it work? flux[x] Big data is…..“ A set of technology advances that have made capturing and analysing data at high scale and speed vastly more efficient” Data scientists R Hadoop [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 50. How does it work? flux[x] “big data” Your data [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 51. How does it work? flux[x] “big data” 6 5 4 3 2 1 0 Your data Category 1 Category 2 Category 3 Category 4 [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 52. flux[x] IN THE REAL WORLD [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 53. In…retail flux[x] Then – have a sale Now – target individuals [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 54. In…online marketing flux[x] Now – triangulate multiple Then - pipeline data sets to identify targets [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 55. In…IT flux[x] Then – data locked in Now – rich, real time unusable formats analytics [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 56. In…risk management flux[x] Now – telematics and Then – paper validation unstructured data [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 57. In…law enforcement flux[x] Now – crime hotspot Then – gut instinct identification [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 58. In…medical research flux[x] Now – semantic and NLP Then – keyword searches “understanding” [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 59. In…sports flux[x] Then – subjective “truth” Now – objective insight [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 60. flux[x] PART 2 IN THE REAL WORLD [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 61. flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 62. flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 63. 90% of colleagues don’t really care flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 64. A Vanguard doesn’t really exist flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 65. Use cases are still limited flux[x] Digital marketing x Data retention, d ata archiving Big Data exploration and discovery x Data x Machine- Fraud generated detection data and xanalytics Social network and relationship analysis x prevention [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 66. Hires are scarce… flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 67. R and Hadoop aren’t for all techies… flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 68. flux[x] Big data only shows us what’s not obvious… [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 69. So…using big data actually means…. flux[x] / Colleagues who are biologically pre-programmed not to care / No best practice to rely on / Very specific use cases within business / Hiring aching expensive people / To use software anyone with an IQ <150 shouldn’t go near / All in the cause of…..looking for non conclusive, obscure patterns, which may or may not lead to any business value [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 70. flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 71. flux[x] PART 3 STRATEGY BY EXPERIMENTATION [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 72. Strategy by experimentation flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 73. 1. Set aside some time (and budget) flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 74. 2. Find a partner you trust flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 75. 3. Get everyone involved flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 76. 4. Identify a business issue you have… flux[x] What kind of Why is the ROI people make the on my CPC largest account purchases? declining? There’s always a How can I drive sales spike on sales of Thursday recommendation evenings – why? s in the check out process? [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 77. 5. Identify data sets which might hold flux[x] clues Adwords data Social data Web analytics data There’s always a sales Store footfall spike on Thursday evenings – why? Brand tracking Transactional data Logistics files [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 78. 6. Play detective flux[x] Patterns Relationships Correlations [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 79. 7. Rinse and repeat flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 80. 8. Re-humanise the story… flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 81. 9. Shout about it flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 82. 10. Enjoy the journey flux[x] [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 83. To summarise flux[x] / Define big data for yourself – see how it could apply to you / Accept the challenges / Start small, grow and publicise success [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 84. flux[x] THANK YOU! [x] fluxx.uk.com/consults © 2012 – All rights reserved Fluxx Ltd.
  • 86. How not to eat an elephant Suranjan Som | Head of BI Practice
  • 88. “As consumer technology innovation and a culture of social decision-making changes the way that our customers make decisions about their purchases, it is becoming increasingly difficult to differentiate our products and services from those of our competition”
  • 91. Customer intelligence has historically relied on assumptions based on transactional behavior analysis
  • 95. “Big Data presents us with some very exciting short and medium term wins around customer insight. But the most strategic opportunity for us is in the integration of Big Data to enrich the intelligence we hold around our single view of customer”
  • 97. Business Transformation & Enablement Organisational Governance Compliance & Regulatory Information Governance Information Delivery Solution Delivery Information Services Information Stores Information Integration Infrastructure
  • 101. Outsource Risk Specialist Business Users Blended Resource Data Scientists Opportunity Data Services & Operations Technology Specialists
  • 102. Data Production Data Storage & Management Data Exploitation 1. Data Acquisition 2. Data Transformation 3. Data Enrichment 4. Data Dissemination 5. Data Disposal 6. Data Governance 7. Data Lifecycle Management
  • 104. Analytics Multi-platform Delivery Document Management Business Abstraction Layer Big Data Relational Data Unstructured Data Structured Data Documents
  • 107. “Big data” is based on a single fear: There might be something powerful and useful in all that data! thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 108. “There are things we know we don’t know” Donald Rumsfeld thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 109. So the holding position has been hang on to it, otherwise we’ll regret it later thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 110. thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 111. thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 112. So the holding position is A total boon for the storage industry! But equally,… it’s hard to argue with! thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 113. But our (real) experts have told us: 1. It doesn’t have to be big to be useful 2. Big doesn’t always mean complex 3. The insight is worth it thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 114. In other words Size isn’t important, it’s what you do with it that counts thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 115. In other words Size isn’t important, it’s what you do with it that counts thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 116. In other words Size isn’t important, it’s what you do with it that counts thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 117. In other words Size isn’t important, it’s what you do with it that counts thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 118. In other words Size important isn’t . Matters only what you do with it. thinks… © 2012 – All rights reserved Fluxx Ltd.
  • 119. thinks… © 2012 – All rights reserved Fluxx Ltd.

Editor's Notes

  1. &lt;graphic of a processing time versus computing power graph that shows how long it used to take to run a report across a series of retail stores and how this has come down over time&gt;&lt;Then another that shows what happens to the time taken when the amount of data goes up by 100-fold&gt;
  2. &lt;graphic of a processing time versus computing power graph that shows how long it used to take to run a report across a series of retail stores and how this has come down over time&gt;&lt;Then another that shows what happens to the time taken when the amount of data goes up by 100-fold&gt;
  3. &lt;graphic of a processing time versus computing power graph that shows how long it used to take to run a report across a series of retail stores and how this has come down over time&gt;&lt;Then another that shows what happens to the time taken when the amount of data goes up by 100-fold&gt;
  4. &lt;graphic of a processing time versus computing power graph that shows how long it used to take to run a report across a series of retail stores and how this has come down over time&gt;&lt;Then another that shows what happens to the time taken when the amount of data goes up by 100-fold&gt;
  5. The secret to analysing and understanding large data sets is to look at them in a different way.The most valuable analytical tools for large data sets are those that can show the data in a way we can understand them more easily.
  6. 574
  7. Just because data volumes are large, this doesn’t mean that all data captured is accurateFacebook now has “1 billion” active visitorsBut it also can automatically add “likes” without that user clicking the like button (or even seeing the page)This raises questions about the accuracy of the data, but also whether all metrics, like visitors, are correct
  8. With the rise of available data, new fields such as behavioural economics are emerging that look to understand what forces can influence customers’ behaviour and decision-making.This means that data can be telling you the right things, even if it doesn’t seem to make sense at first sight
  9. Traditional MI “cube” databases are very good at regular reporting But they are less good at running detailed queries against massive data sets for initial researchLinear databases are much more efficient for initial researchConsider using two tools to analyse dataOne for regular reporting, and a “rapid response unit”Linear databases can give you different answers to what appears to be the same questionBut they can also answer questions that most MI tools cannotThis means it is really important to understand both the tools you have and the data you hold to ensure that you are answering them correctly
  10. Give them the tools to find the insights that you are looking for.Be prepared to use different tools to analyse large data sets
  11. If the ultimate concept of celebrity, or at least popularity, is politics, then Obama’s advert for data analysts shows how valuable data and its correct understanding can be.
  12. A global telco’s call centre in the US wanted to analyse the multichannel data that influenced usage of their call centre.This included web traffic that preceded and followed a call to the call centre, and what insights could be gained and recommendations to drive improvements.Initial analysis to identify 10 optimisation recommendations took less than one week.By reducing call centre traffic by 2%, the telco established it would save $600m per annum.
  13. break it down, Demystify it
  14. Marketing and intelligence point of view
  15. New products, new channels to market, new customers
  16. Getting to that insight is exponwntiallyahrder than everyone says it is, getting that insight used and understood is even harderWhy?