atme travel marketing conference - how big data, deep web & semantic technologies change travel...
TRANSCRIPT
How Big Data, Deep Web & Semantic Technologies Change Travel Marketing
ATME Marketing Conference Hyatt Regency – Miami, Florida April 17, 2013
Image Credit: NASA Goddard Photo and Video (cc|flickr)
Session Overview
• Big Technology – Big Data – The Deep Web – The Semantic Web
• Big Impact – Curing What Ails You – You Are Already Using Big Data – Travel Examples
• What Could Possibly Go Wrong? • The Future
– Big Travel – The Killer App
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 1 Image Credit: NASA Goddard Photo and Video (cc|flickr)
Big Data IS a Big Deal
• What is Big Data? – Too Much, Too Fast, or Too Weird to Fit in a Database
• Why is There Big Data? – Web 2.0 – Introduced User Generated & Shared Content
• How Big is Big? – US Library of Congress = 235 Terabytes Data
– 30 Billion Pieces of Content Shared/Month on Facebook
– $600 Disk Drive Now Stores All The World’s Music
• You’ve Already Seen Big Data at Work – Google Analytics – Page Traffic, Source & Navigation
– Netflix – Recommendations & Manages its Video Streams
– Jeopardy – IBM’s Watson Computer Beat Ken Jennings
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 2 Statistics: McKinsey Global Institute 2011 | Image Credit: NASA Goddard Photo and Video (cc|flickr)
Big Data – Big Insights
• Definition – “Big Data is High Volume, High Variety, High Velocity, High
Veracity Information Assets for Enhanced Insight and Decision Making”
• High Volume – 90% of World’s Data Created Over Last Two Years
• High Variety – Structured | Unstructured | Video | Images | Signals
• High Velocity – Real-time Response | Engage Customer | Avoid Fraud
• High Veracity – 1/3 Business Leaders Don’t Trust Information for Decisions
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 3
Keys to Big Data Success
• Robust Platform
– Good Technology (Hadoop / MapReduce)
• Accurate Algorithms
– Good Math
• Smart Insights
– Good Interpretation
• Effective Communication
– Correct Context & Relevance
• Solution: The Right People
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 4
Major Big Data Challenges
• Human Capital Shortage
– US Needs 140,000 to 190,000 More People with Deep Analytical Skills
– US Needs 1.5 Million Managers & Analysts to Analyze Big Data to Make Better Decisions
• Adopting New Technological Infrastructure
– Navigating a Transitional Roadmap
– Purging Inaccurate Data from Existing Systems
• Privacy Safeguards
– Protecting Consumer & Corporate Data 2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 5 Statistics: McKinsey Global Institute 2011 | Image Credit: NASA's Marshall Space Flight Center (cc|flickr)
Google Versus Swine Flu
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 6 Image Credit: www.brettarthurphoto.com (cc|flickr)
• Predictive Analytics
– Discovered Close Relationship Between Web Searches for Flu-related Topics & People with Flu Symptoms
Target Versus Pregnancy
• Opportunities to Change Shopping Patterns
– Graduation | Relocation | Job Change |Child Birth
– Traditional Method: Source Birth Records
– Shift Change: Target Identifies in 2nd Trimester
• Predictive Analytics
– Identify Conscious & Unconscious Patterns
– Establish Cue-Routine-Reward Loops
– Email Coupon – Weekend Shopping – Free Starbucks
• Every Purchase Linked to Guest ID
– Look at 25 Products Together – Estimate Delivery Date
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 7 Image Credit: MOCO Big Ideas Blog
Jack Andraka Versus Cancer
• 15 Year Old Kid – Limited Access to Unstructured Big Data
– Research Sources – Google & Wikipedia
– Early Detection Test for Pancreatic / Lung / Ovarian Cancer
• Diligent Effort + Innovative Idea = Massive Disruption – 8,000 Possible Proteins – Identified Mesothelin Biomarker
– Idea: Carbon Nanotubes Bind Antibodies/Electrical Charge
– Sent 200 Lab Space Requests; Received 199 Rejections
• A Faster / Better / Cheaper Test – Takes 5 Minutes versus 14 Hours
– 400x More Sensitive & 50% More Accurate
– $0.03 Per Paper Strip (10 Tests per Strip) versus $800/Test
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 8 Image Credit: TED Conference (cc|flickr)
Keys to Big Data Success
• Base on Business Needs, Not Tech Dreams – Executive Level Support
– Make Data Driven Decisions
• Align Projects with Organizational Goals – Big Data Initiatives to Support Specific Objectives
• Start Small – Use Early Successes to Demonstrate Benefits
– Gain Momentum
• Expand from Foundational Projects – Extend Capabilities By Adding New Data Sources
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 9 Image Credit: Antique Hardware (cc|flickr)
The Deep Web
• Definition:
– Underlying Data that Populates Dynamic Web Pages and is Not Captured by Search Engines
• Airline Example:
– Airfares, Seat Inventory, Fare Rules, Bag Fees
• Hotel Example:
– Stay, Arrival, Extra Guest, Room Type, Bed Type
• Car Rental Example:
– Weekend, Time, One Way, Discount Eligibility
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 10 Image Credit: g-na (cc|flickr)
Website Email Offers
Upsell / Cross-Sell
Search
Advertising
Social Media
Advertising
Location Services
Ratings & Reviews
Virtual Reality
CRM / Loyalty
Counter / On-board Call Center
Kiosks
User Platform
Wallet
Form of Payment
Payment Processor
Tech Provider
Direct Connect
GDS
Inter-mediary
Supplier
Partner
Structured
Unstructured Multimedia
The Growing Abyss Of Travel Data
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 11 Content: OpenTravel Alliance / RockCheetah
Why Google Bought ITA Software
• Access to Airline Deep Web Data
• Search from Hell
– 5 Origin Airports
– 5 Destinations
– Different Dates
• Crazy Fast
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 12
Semantic Web – Provides Context
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 13
• Before: The Internet of Pages
– Search Engines Return Pages with Lots of Content
– Mixed Page Content - Varying Degree of Relevance
• Now: The Internet of Data
– Profile & Navigation Determines Dynamic Content
– Travel Industry Largely Lacking Semantic Schemas
• Next: The Internet of Things
– 25 Billion Connected Devices | 2015
– 50 Billion Connected Devices | 2020
Image Credit: ...-Wink-... (cc|flickr)
Why Context is Important:
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 14 Image Credits: Tambako the Jaguar, RobDurfee &mutrock (cc|flickr)
Jaguars
The Semantic Web | 2 + 2 =5
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 15
• Definition: – Association Indicating How One Entity Relates to One
or More Other Entities to Create Meaning that Goes Beyond the Components Themselves
• Key Benefit – Allows Systems to Understand Relationships
• RDF Triple - Resource Description Framework – Simple Subject | Predicate | Object Structure
• Examples: – I Know Henry | Eric Clapton Plays Guitar
– Structure Applies to EVERYTHING
Image Credit: GustavoG (cc|flickr)
A Common Vocabulary is Essential
• GoodRelations Ontology
– Yahoo!
– Best Buy
– Bing
– Volkswagen
• Requires Cooperation
– Travel Suppliers
– Technology Providers
– Intermediaries
– Retailers
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 16 Image Credits: Millzero Photography (cc|flickr)
GoodRelations Vehicle Rental Vocabulary for Car Descriptions
• vso:ACRISSCode
• vso:VIN
• vso:acceleration
• vso:axles
• vso:bodyStyle
• vso:cargoVolume
• vso:color
• vso:condition
• vso:damages
• vso:doors
• vso:driveWheelConfiguration
• vso:engineDisplacement
• vso:engineName
• vso:enginePower
• vso:engineType
• vso:feature
• vso:firstRegistration
• vso:fuelConsumption
• vso:fuelEconomy
• vso:fuelTankVolume
• vso:fuelType
• vso:gearsTotal
• vso:height
• vso:length
• vso:meetsEmissionStandard
• vso:mileageFromOdometer
• vso:modelDate
• vso:payload
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 17
• vso:Automobile (rdf:type owl:Class) - - - - - - - - - - - - - - - - - - - - Defines a Classification Named “Automobile”
• rdfs:subClassOf vso:MotorizedRoadVehicle - - - - - - - - - - - - - - - Automobiles are a type of Motorized Vehicle
• URI http://purl.org/vso/ns#Automobile - - - - - - - - - - - - - - - - - Resource For Car Rental Relationships
• vso:previousOwners
• vso:productionDate
• vso:rentalUsage
• vso:roofLoad
• vso:seatingCapacity
• vso:speed
• vso:steeringPosition
• vso:tongueWeight
• vso:trailerWeight
• vso:transmission
• vso:weight
• vso:weightTotal
• vso:wheelbase
• vso:width
Image Credit: antonychammond (cc|flickr)
Graphic Source: Tim Berners Lee & World Wide Web Consortium (W3C) | Image Credit: MoHotta18 (cc|flickr)
The Semantic Stack - Technology
• A Lot More Technologies Involved
• Address
– Vocabulary
– Search Queries
– Rules
– Validation
– Security
• Some Don’t Exist
– Proof / Trust
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 18
User Interface & Applications
Data Interchange: RDF / RDFa
Unifying Logic
Encryp
tion
Proof
Rules: RIF
Ontology: OWL
Taxonomies: RDFS
Query: SPARQL
Trust
Syntax: XML / XML Namespaces
Character Set: Unicode Identifiers: URI
Au
then
tication
Who’s Investing in Semantic Web?
• Schema.org – Semantic Search
– Bing, Google, Yahoo! & Yandex Cooperating
• Google Knowledge Graph - Immediate Answers
– Integration with Google Places | Local | Google+
– Mobile Integration in Android Google Now Cards
• Facebook Open Graph - Apps Tell Stories
– About People
– The Things They Did
– Who They Were With
– Places Where It Happened
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 19
TripAdvisor & Airbnb v. Review Spam
• Both Utilize the Facebook Open Graph – Simplified Registration Process
– Filter Results For Connected Listings (Airbnb)
• Benefits – Social Validity – Opinions of Real People You Know
– Social Authority – Identify Thought Leadership
– Social Commonality – Identify Similar Preferences
• Results – Average User Engagement Increased 20% (TA)
– Connected Facebook Users Contribute Content 2x More
– Gained Access to Primary Email Address (RKC)
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 20
Image Credit: gato-gato-gato (cc|flickr)
Keys to Semantic Web Success
• Start with Consumer Marketing Needs – What Type of Content is Desired?
– How is it Consumed?
– What Questions Must Be Answered?
• Identify Available Data Sets – Start with the Linked Data Cloud
• Results from Using Sematic Structure – Best Buy: 30% Increase in Store Page Traffic
– Yahoo: 15% Increase in Click Through Rates
– Volkswagen UK: Easily Integrates 3rd Party Content
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 21
Image Credit: alvaro.stuardo (cc|flickr)
What Could Possibly Go
Wrong?
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 22 Image Credit: Public Domain
Bad Data - The Four Seasons Bathrobe
• Mid-1980’s – The Dawn of Word Processing
• Background – Hotel 60% Repeat Guest Ratio
• The Challenge – Induce 1st Timers to Return
• Poorly Executed Cue-Routine-Reward
– Send 1st Time Guests Complimentary Bathrobe
• Problems
1. Poor Matching – Sent to Frequent Guest
2. Sent to Guest Home Address – Surprise!
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 23
Innovative Technology on Back Burner
• Sabre Labs – Application Prototypes (1998)
– Collaborative Filtering for Hotel Recommendations
– Drive-pathing for Auto Travel Itineraries
• SideStep.com – Meta-search App (2004)
– Acquired by Kayak | Site & App Shut Down
• UpTake.com – Semantic Reviews (2008)
– Acquired by Groupon | Site Closed
• Room 77 – Hotel Room-specific Reviews (2011)
– Hotel Systems Could Not Support Functionality
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 24
New Distribution or Old Guard? • ATPCO Owners
– OneWorld • AA-BA-IB-JL
– SkyTeam • AF-DL-KL
– Star Alliance • AC-LH-SK-SR-UA-US
– Others • AS-FX-HA
• OpenAxis Founders
– OneWorld • American Airlines (AA-US)
– Sky Team • Delta Airlines (DL)
– Star Alliance • Air Canada (AC) • United Airlines (UA-CO)
– ATPCO (Allied) – Farelogix (Allied) Schema
• IATA Board of Governors – OneWorld
• BA-CX-IB-JL-LA-QF
– SkyTeam • SU-AM-AF-MU-CZ-DL-KQ-KL-KE-SV
– Star Alliance • AC-MS-LH-SR-TK
– Others • KM-AV-ET-EY-FX-GA-9W-B6-QR
• GDS Founders
– Amadeus (AF-LH-IB-SK) – Sabre (AA) – TravelPort
• Apollo (UA) • Galileo (AC-BA-KL-AL-AZ-SR-OS-OA-
SN-TP-EI) • Worldspan (DL)
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 25 Image Credit: Richard Cawood (cc|flickr)
Obstacles That Postpone Disruption
• Three Horseman of Stagnation – Control – Dominant Partners in Business Model
– Capital – If Allocated to Maintaining Status Quo
– Captive Consumers – Significant Barriers to Entry
• Focus Becomes Company – Not Customer – Defend Position
– Manage to Quarterly Results
– Core Business Orientation – Diversification Resistance
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 26 Image Credit: Adam Foster | Codefor (cc|Flickr)
The Future – Major Tech Disruption
• New Interfaces – Google Glass Visual Display
– Apple Siri Voice Input
• New Payments – Square | Wallet Provides Seamless Payments
– Dwolla | $0.25 Fee for Any Purchase Amount
• Self-Driving Cars – Google / Toyota / Texas Instruments | Target 5 Years
• DNA Storage – 700 Terabytes of Data Stored in Single Gram of DNA
– Dense | Volumetric | Stable (Thousands of Years)
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 27 Image Credit: kevin dooley (cc|flickr)
Marketing Challenge: Mouths to Feed
• Suppliers – Must Manage Complex Operations
• Technology Providers – Design & Support Innovative, Stable Platforms
• Wholesalers – Aggregate Demand Better Than Single Suppliers
• Retailers – Understand Merchandising & The Customer
• Consumers – Want Value & Best Available Deal
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 28 Image Credit: mclcbooks (cc|flickr)
Big Travel: Seven Step Travel Process
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 29
Inspiration
Research
Planning
Validation Booking
Travel
Sharing
Image Credit: Andrew and Annemarie (cc|flickr)
Big Travel: Five Travel Transactions
• Stand-alone Component – Independent Purchase
• Value Added Combination – Hotel Room with A Meal, Massage or Round of Golf
• Bundled Purchase – Multiple Interchangeable Travel Components
• Dynamic Packaging – Dynamically Priced / Rules Based / Collaborative Process
• Sequential Packaging – Purchase Enables Secondary Offer or Product Multiples
• Distributed Packaging – You’ll Need to Ask Me About This & Sign an NDA
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 30
• Home to Airport (Taxi / Limo / Car & Parking)
• Airport Experience (Waiting & Boarding)
• In-Flight Services
• Ground Transfer to Hotel | Car Rental
• Hotel | Meals | Tours | Activities | Memories
• Hotel to Airport Transfer | Car Return
• Return Flight
• Airport to Home (Taxi / Limo / Car & Parking)
Big Travel: End to End Experience
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 31
Big Travel: Big Personalization
• Traveler Personas
– May Be Travel Agent or Supplier Brand Loyalty
– May Prefer Packages or Avoid Packages
– May Seek Same or Different Destinations
• The Multi-Persona Traveler
– Needs Change Based on Itinerary
– Business Trip | Family Vacation | Spouse Getaway
• Multiple Travelers Create Complex Itineraries
– Couples & Families with Divergent Interests
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 32
The Killer Travel App - Product
• Provides the RIGHT Experience
• To the RIGHT People
• Doing the RIGHT Things
• In the RIGHT Places
• At the RIGHT Times
• Through the RIGHT Channels
• With the RIGHT Products
• At the RIGHT Price
• And the RIGHT Value
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 33 Image Credit: Tom.Bricker (cc|flickr)
The Killer Travel App - Technology
• Accurate
– Based on the Latest Real-time Information
– Not a List of Options | A Suggested “BEST” Option
• Personal
– Relevant to Specific Needs of That Traveler
– Customized for That Particular Itinerary
• Fast
– 1-second Delay = -11% Pageviews, -7% conversions & -16% Customer Satisfaction
• If Really Good, Gives Answer Before Being Asked
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 34 Statistics: Aberdeen Group , 2011
The Killer Travel App - Efficiency
• Leverages Industry Standards – Facilitate Communication
– Streamline Processes
• Technology Adds Functionality / Reduces Cost – Big Data
– Deep Web
– Semantic Relationships
• Disruptive Innovation – A Holy War to Democratize Information
– Some People Are Not In It for the Money
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 35
Multi-Origin / Multi-Destination Multi-Modal / Personalized Itinerary
5-Apr 6-Apr 7-Apr 8-Apr 9-Apr 10-Apr 11-Apr 12-Apr 13-Apr 14-Apr 15-Apr 16-Apr
Thu Fri Sat Sun Mon Tue Wed Thu Fri Sat Sun Mon
MKE-ORD
Avistar ORD
LH ORD-MUC LH MUC-CDG TP SVQ-LIS
TP LIS-LHR
U2 VCE-CDG UA LHR-ORD
SNCF
Hotel Windsor-Opera (PAR)
Versailles
Notre Dame
Louvre
Eiffel Tower
d'Orsay
PAR-MAD
MAD-SVQ
Hotel Belquer (SVQ)
Giralda
Carriage
Alcazar
ORD-MKE
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 36
Impossible? Remember Jack Andraka?
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 37
Embrace Big Data, Deep Web & Semantic Web
2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 38
• Or Don’t…
• Millions of Kids, Like Jack, Will Be Disruptive
• Smartly Using Available Technology
– Better
– Faster
– Cheaper
• Maybe Without the Same Profit Motive
How Big Data, Deep Web & Semantic Technologies Change Travel Marketing
Questions?
Copies of the Presentation? Comments?
Robert Cole skype: robertkcole twitter: @robertkcole phone: +1.262.309.9560 web: rockcheetah.com email: [email protected] 2013 ATME Marketing Conference Big Data, Deep Web & Semantic Technologies Robert Cole ● RockCheetah | 39