Master big data entreprise, business model & individuals

BIG DATABusiness models

& Platform economy

April 2016

Yves Eychenne

AGENDA• Big Data overview

• Big Data, uses cases and impact on Business models

• Big Data: how to manage privacy and user acceptance?

THE ERA OF INFORMATION

Datafication

Ubiquity

Internet of Things

Social networks

Census, Polls

… maritime routes.

Origins of BIG DATA

Big Data definition: the 4 V

VOLUME VELOCITY VARIETY VERACITY

DATA AT REST

Terabytes to Exabytesof data to store and

process

DATA IN MOTION

Streaming data, in seconds or milliseconds,

to take action

ANY DATA

Structured, unstructured, text,

multimedia.

DATA IN DOUBT

Uncertainty, inconsistency, incompleteness, latency: the results will be a probability

VOLUME

Number of data elements to capture

All the data, no sampling

Structured

Semi – structured

Un structured

Variety

Velocity

Real time

Stream computing

Inconsistency

Incomplete

Ambiguity

Latency

Measurement

Veracity

BIG DATA – TOOLSData Storage (Hadoop) cheap, supports parallel processing.

Machine learning to recognize models or patterns.

Predictive analytics to find problems before they occur.

Stream processing to detect patterns in real time, to correlate data produced in real time vs historical data.

Visualization tools

Source IBM – Visualization samples

CORE ENGINE OF BIG DATA :

MACHINE LEARNINGA legacy of Artificial IntelligenceMachine learning can match a large amount of data in input

It learns, in self sufficient and adaptive mode, to classify and process data, after a phase where the tool learns with a Data Scientist (like a kid with its parents)

The processing of the data will be done with a level of probability

A feedback loop helps the tool to get better over time.

Grammar corrector (Microsoft, Google.)

Hand writing recognition

Kinect of Microsoft.

Medical diagnostics

GPS navigation (Waze)

Financial market analysis

Predict failure of industrial equipment

Automated indexation and searches of images and videos

Humanoid robots (movement control)

MACHINE LEARNING USE CASES

Visualisation

New tools

FROM BIG DATA TO COGNITIVE SYSTEM: IBM WATSON

Extraction models

EXPLORE THE DATA

Trouve les modèles

Action !

Use models to trigger action

1. Natural LanguageImage, voice, emotion

2. Evaluatehypothesis

3. The system learns and improves

FIND THEMODELS

ANALYTICS

AGENDA• Big Data overview

• Big Data, uses cases and impact on Business models

• Big Data: how to manage privacy and user acceptance?

WHAT BENEFITS FOR COMPANIES ?

From consumer to consum’actor

The consumer in the era of information is a consum’actorconnected and informed.

Businesses and brands should:

Adopt new business models

Listen to the consum’actor through social network

Understand, offer new experiences, capture trends and disruption.Be relevant for the new generations (X, Y)

Rolls Royce does not sell engines any more by Total Care service (“Power by the hour”)

From service to experience (smartphone, Google, etc.)

Share economy 2.0 (BlablaCar, Uber, AirBnB)

PRODUCT, SERVICE, EXPERIENCE

https://datafloq.com/read/rolls-royce-shifts-higher-gear-big-data/514

A REVOLUTION ? DIGITAL GIANTS vs TRADIONAL ECONOMY

The importance of a new busines model….New Economy Region 2016Market B=Value ($B) Traditional Economy

equivalentRegions 2016

Apple USA 583 Sony (electronic) Japan 45

GoogleFacebookBaidu

USAUSA

China

516340

Publicis (pub) USA 13

MicrosoftSaleforce

USAUSA

43950

Oracle (Software) USA 170

AmazonAlibabaPricelineeBay

USAChinaUSAUSA

2811926527

Walmart United States 221

Yahoo! USA 36 News Corp AUS 7.6

UBER USA 40 Medallion CorpTaxi G7

USA 220M

LinkedIn USA 15 Manpower USA 6

Netflix USA 45 Vivendi France 24

2004 data as of 9/17/2004. 2013 Market value as of 12/19/2013. 2012 revenue is TTM.

List excludes Alibaba ($75B), whose private market value would put it in the Top 10.

List also excludes Skype (bought by MSFT in 2011 for $8,5B), YouTube (reported as part of Google) and Paypal (reported as part of eBay)

Exercise: the 2000 Internet crash (1 points)

• Study Yahoo capitalisation in 1999/200 through Tom von Altencomments on his blog

• http://fortboise.org/momentum.html

• Questions: what was wrong in Yahoo capitalization …. And business model?

• Hint: The audience was the key to determine the market cap…. Whatwas missing?

TWO SIDED BUSINESS MODEL

AND THE PLATFORM REVOLUTION« To succeed, platforms in industries such as software, portals and media, payment systems and the Internet, must get both sidesof the market on board. Accordingly, platforms devote much attention to their business model, that is how they court each sidewhile making money overall. » Jean Tirole, Jean Charles Rochet, « Platform Competition in Two-Sided Markets » University of Toulouse report, Dec. 3th 2002.

Platform revolution by Sangeet Paul Choudary, Geoffrey Parker and Marshall W. Van Alstyne, 2016.

http://platformed.info/best-of-platformthinking-2015/

BIG DATA, THE ENGINE OF THIS REVOLUTION

Analytics and big data enable new business models :

Platform /Two sided business models (Google, Yahoo!, Apple).

Freemium (LinkedIn, Spotify, Deezer).

Social selling (Amazon, Booking, Tripadvisor, Netflix, eBay, LeBonCoin).

Share economy (Uber, BlablaCar, AirBnB).

… the importance of customer experience and being relevant!

And Small & Medium Business?

SMB can buy services leveraging Big Data :

Digital marketing / ads

Social network sentiment analysis for their products, services & brand

Predictive maintenance

Specialized services Drone & Big Data for the agriculture

SMB (French market ) can also create services on top of Big Data :

CRITEO ($2.3B NASDAQ).

Exalead (bought by Dassault Systems).SMB can propose new services with Big Data :Withings : connected health and Big Data.

Retency (Big Data for retail store).

Crazylog. Predictive maintenance service

NEW BUSINESS MODELS AND NEW VALUE PROPOSITION

Do you know how to model?

Two sided markets

Freemium services

Sharing economy

New supply chain model

Customer experience (mobile, social).

http://www.businessmodelgeneration.com/

Google: the importance of analytics on a leading multi-sided business platform

Monetizing SEARCH

Banner ads (1994)

Licensed search (1996)Alta Vista to Yahoo

Fixed rate keyword listings (1997)

Keyword auction (2000)

Overture: pay for performance

Enhanced keyword auction (2002)

Google’s AdWords:

https://adwords.google.com/select/Login

“Contextual” based listings (2003)

Google’s AdSense, Ad Exchange

https://www.google.com/adsense/

• Many search engines, only one winner – why?• A good search engine

• Also business model: find the right value set of keywords

• … show to customers and get them to pay again and again

• This requires big data analytics. Google was among the inventors of that concept.

Other examples of analytics in multi sided platforms

• Facebook

• Advertising on PC and Mobile = > Brand « presence » based on you profile

• Advanced ad auction platform• https://www.facebook.com/help/163066663757985

• Facebook insights

• Location service in the future (CISCO+Facebook: free wifi service in US)

• http://www.engadget.com/2013/10/02/facebook-cisco-free-wifi-for-checkin/

• Telco are trying to play the same game

• Linkedin

• Advertising, auction platform

• Premium services

• Recruiting platform (matches profiles)

• Apple: pay on both side on the platform

• The device, the Apps, music

• Some apps and services are free

• They use their brand image: luxury segment

• Shopping Mall

• Signage management

• Site analytics: Ex http://www.footfall.com/site-analytics-property/

• Localization services, sms promotion

• Real Estate

• New model

• Match offer – work for both side

• Bring service for a lower fixed price to segmented customer

FREEMIUM SERVICES

• Many Games on smartphone and tablet

• Newspapers on tablets

• Google maps• Initial 1000 api call are free then

you have to pay

• Analytics support for Freemium:• Free / paid customer balance

conversion

• Conversion from free -> paid service

• Analytics is part of the product (Google API)

• Retention / churn analytics

FROM PRODUCT TO SERVICES:Rolls Royce Aero EnginesTotal Care service (“Power by the hour”)

In 1996, CEO John Rose decided to forward integrate into providing ‘service’

• Service was licensed out to independent firms

• Rationale was that service was more profitable than OEM

• Service covered both regular maintenance and unscheduled repairs

RR charges customers (Airline) on basis of number of hours flown (“Power by the hour”)

• Provided airline with ‘fixed per hour cost’

• Provided RR a clear incentive to minimize service time

Today, ‘service’ accounts for around 50% of RR revenues

• RR developed the means to monitor engines in flight (“Engine health monitoring”)

• Analytics: predictive maintenance and optimized supply chain• Identify when maintenance needed • Identify problems at early stage so

that repair can be optimized

• Analytics for asset management, spares logistics, portals for engine-related information ….

DIGITAL ADVERTISEMENT(37B$ market in US), powered by BigData

A Tour of Online Display Advertising; DSP, DMP, RTB, Ad ExchangesPublished October 13, 2012 | By Paul Mosensonhttp://www.nusparkmarketing.com/2012/10/a-tour-of-online-display-advertising-dsp-dmp-rtb-ad-exchanges/

Software defined supply chain, the 3D printer platform?

IBM “Three emerging and converging technology revolutions - 3D printing, intelligent robotics and open source electronics - are transforming the global supply chain.

• 3D printing reverses standardization

• Intelligent robotics reverses modularization

• Open source electronics accelerates digitization

• Is it real? Eric Currell ‘s success stories• Withings

• Sculteo

• Orange sells iphone case 3D printed

• http://3dpcase.sculpteo.com/en/

This is about the long tail analytics: you can build smallseries targeted to small customer segments

http://sloanreview.mit.edu/article/with-3-d-printing-the-shoe-really-fits/Nike and New Balance started to use 3D printing for Shoes

Paul Mulzoff | May 28, 2013 I believe this is just another example of the evolution of data analytics being used to deliver a more targeted product to consumers. The need for producers to transform the needs of consumers into a more customized product will in turn push the evolution of mass production into more industries. …”

Social selling: analytics is at the core

• Amazon, eBay• Salesforce• Workday• IBM Commerce Platform

The consultant corner:• System of Engagementhttp://www.forbes.com/sites/joshbersin/2012/08/16/the-move-from-systems-of-record-to-systems-of-engagement/

• Gartner: Nexus of Forcehttp://www.gartner.com/technology/research/nexus-of-forces/

The business model canvas

http://www.businessmodelgeneration.com/ HOW IT WORKS: https://www.youtube.com/watch?v=QoAOzMTLP5s#t=27

http://bmimatters.com/tag/business-model-canvas-examples/

http://bmimatters.com/tag/business-model-canvas-examples/

BIG DATA IMPACTS MANY ASPECTS OF THE NEW BUSINESS

MODELS

ManagementOf

Partners & suppliers

E-REPUTATION and BRAND

MANAGEENT OF SALES, INVENTORY AND DYNAMIC PRICING

Big data for e-commerce and stores

Customer analysis

PERFORMANCE of Digital AD

Customer segmentsProduct evolution

Product maintenance,

Exercise 2: define the value of big data withinthe business model canvas (1/2) (3 points)• Teams of three people

• Duration 30 min

• Fill in the business model canvas

where big data can provide value

and explain why it brings value

http://www.businessmodelgeneration.com/

Exercise: define the value of big data withinthe business model canvas(2/2)• Team 1: Auto industry

• Hint: think Telsa, Autolib, new services for connected car

• Team 2: Connected watchmanufacturer

• Hint Apple watch competitor

• Team 3: Farm (yes a Farm!)

• Hint: how can I use drones?

• Team 4: Sharing economy app

• Hint: airbnb, blablacar, etc

• Team 5: Bank industry

• Hint: New payment model, paylib, paypal, new financing model (micro loans, …)

• Team 6: Luxury company

• Hint: going from exceptional and unique product to exceptional service

AGENDA• Big Data overview

• Big Data, uses cases and impact on Business models

• Big Data: how to manage privacy and user acceptance?

BIG DATA And INDIVIDUALS?

Employees

Consumers

Citizens

Data Privacy

Big data creates certain level of reaction and emotion: the « Big Brother » syndrome. Its usage by companies must have some limits and must be beneficial to both the enterprise and its customers & prospects.

Two thirds of people are willing to share some information but on the condition of getting something in return.

Waze (GPS service of Google) collects your GPS data but in exchange you get an updated map and accurate traffic and directions, (in addition to some ads).

The relationship must be balanced. The enterprise must collect only the required data for the service (and not all your data) and must be transparent about its usage.

The LIMITS on PERSONAL DATA

(PRIVACY)«Constitue une donnée à caractère personnel toute information relative à une personne physique identifiée ou qui peut être identifiée, directement ou indirectement, par référence à un numéro d’identification ou à un ou plusieurs éléments qui lui sont propres. Pour déterminer si une personne est identifiable, il convient de considérer l’ensemble des moyens en vue de permettre son identification dont dispose ou auxquels peut avoir accès le responsable du traitement ou toute autre personne.

Constitue un traitement de données à caractère personnel toute opération ou tout ensemble d’opérations portant sur de telles données, quel que soit le procédé utilisé, et notamment la collecte, l’enregistrement, l’organisation, la conservation, l’adaptation ou la modification, l’extraction, la consultation, l’utilisation, la communication ...»

“Constitutes personal data any information relating to an individual identified or can be identified, directly or indirectly, by reference to an identification number or to one or more of its own. To determine whether a person is identifiable, all means should be considered to allow identification available or may have access to the data.

Is a personal data processing any operation or set of operations on such data, regardless of the method used, including the collection, recording, organization, storage, adaptation or alteration, retrieval, consultation, use, disclosure….”

HAVAS MEDIA

FRENCH PEOPLE AND THEIR PERSONNAL DATA (SEPT 2014)

http://www.havasmediaopendata.com

(PEOPLE WILLING TO SHARE THEIR DATA)

HAVAS MEDIA

FRENCH PEOPLE AND THEIR PERSONNAL DATA (SEPT 2014)

http://www.havasmediaopendata.com

(THEIR FEARS)

HOW TO PROTECT YOURSELF ?

The consumer (or the employee or the citizen) needs to learn to manage and to protect his digital identity.

Education and explanation are important to keeping its trust when an company or a state is using Big Data

Enterprises must keep a high level of security and manage the risk associated the use

of Big Data. Strong data anonymisation is also a requirement.

Exercise 3: data privacy (2 points)

•What personal data are you willing to share?

•Do you think that it is possible to anonymize GPS data?

•How much is your data worth?

THE BIG DATA ….. REVOLUTION

…OR EVOLUTION !

Additional reading• Yves Eychenne, Jean-Charles Cointot, “La Révolution Big Data

(Dunod Editor, Strategy and management collection) – published in September 2014

• Viktor Mayer-Schönberger et Kenneth Cukier. Big Data « A revolution that will transform how we live, work and think. Edition : Eamon Dolan Book

• Etude de McKinsey Global Institut : « Big Data : the next frontier for innovation, competition and productivity ». 2011. http://www.mckinsey.com/insights/business_technology/big_data_the_next_frontier_for_innovation

• Gartner, Digital Marketing Spending Survey, 2013 http://www. gartner.com/technology/research/digitalmarketing/digitalmarketingspendreport.jsp

• Lumascape de Lumapartners http://www.lumapartners.com/ lumascapes/displayadtechlumascape/

• Zang Hui, Jean BoLot, “Anonymization of Loca tion data Does Not Work : A LargeScale Measurement Study”, Technicolor, Pro-ceeding MobiCom ‘11 Proceedings of the 17th annual inter na -tional conference on Mobile computing and networking

• Platform revolution by Sangeet Paul Choudary, Geoffrey Parker and Marshall W. Van Alstyne, 2016.

• Business Model Generation, Yves Pigneur et all, Wiley

Additionnal learning

MOOC www.bigdatauniversity.com

Business model applied to Googlehttps://www.youtube.com/watch?v=RzkdJiax6Tw

Value proposition added to business model canvashttps://www.youtube.com/watch?v=aN36EcTE54Q