the platform revolution and healthcare · geoffrey parker professor, thayer school engineering,...
TRANSCRIPT
Geoffrey ParkerProfessor, Thayer School Engineering, Dartmouth College
Director, Master of Engineering Management Program
Research Fellow, MIT Initiative on the Digital Economy
The Platform Revolution and HealthcareScaling digital health innovation through platformsLondon, June 18, 2018
Geoffrey ParkerDartmouth College
@g2parker
Marshall Van AlstyneBoston University
@InfoEcon
Platform Revolution: Making Networked Markets Work for You
Questrom School of Business
2016 Parker & Van Alstyne with Choudary –
licensed under Creative Commons Attribution-
ShareAlike 4.0 Int’l (CC BY-SA 4.0).
Sangeet ChoudaryPlatformation Labs
@sanguit
3
Source: P. Evans, “Networks, Data and Platforms,” in Growing Global: Lessons for the New Enterprise, Center for Global Enterprise, 2015.
Forces of Change – Likely to Intensify
Surge in data and tools that can manage and
analyze data
Networks connect physical, digital, and social
Age of Networks Age of DataFIRM
Age of PlatformsNew business models that that
leverage networks and intelligence
World’s top 50 public platforms by market cap
Source: Data from Thomson Reuters Eikon, May 2017
Supermajor Platforms7 companies = $3.5 trillion
Other Platforms43 companies = $1.4 trillion
Platform Market Caps Displace Energy and Banking
5Source: Visualcapitalist.com, Bloomberg
Many startups use platform business models
… many product/service companies are trying to or already have become Platforms
… all firms will have to deal with platforms
Traditional Linear Chain: “Pipes”
Stage 1 Stage 2 Stage 3
$ $ $
Value accumulates from stage to stageMinimal Network Effects
Structure of Platforms
PLATFORM
Side “B”Side “A”
DirectExchange of Value
Simple at first, then can Become Multisided
Users
3rd-party sites (AdSense)
Advertisers
DirectExchange of Value
Simple at first, then can Become Multisided
Users
3rd-party sites (AdSense)
Advertisers
DirectExchange of Value
DirectExchange of Value
Adapted from: Andrei Hagiu
Platforms tend to get more complex over time
Professionals App Developers
Advertisers Recruiters
Corporations Education
Adapted from: Andrei Hagiu
Same Firm can be both Platform and Pipe
Key difference: cost structure (control & responsibility), risk
SELLERS BUYERS
BUYERSSELLERS
Amazon MarketPlace as Platform
Amazon as Pipe
Adapted from: Andrei Hagiu
BMWBASF
AppleSamsung
AirbnbUber
Platforms Scale Faster than Pipes
Asset LightPlatform
Platform &Pipe
Pipe
A PLATFORM: • Is a nexus of rules and
architecture
• Is open, allowing regulated participation
• Actively promotes (positive) interactions among different partners
PLATFORM
SIDE BSIDE A
Source: Platform Revolution
Firms are investing in building their data layers
17
New locus of value creation and capture
Agriculture
Physical Layer
Energy
Physical Layer
Healthcare
Physical Layer
Banking
Physical Layer
DATA Layer DATA Layer DATA Layer DATA Layer
Source: P. Evans, CGE, 2015
Service value of industrial machines
Compressor Discharge
16th stage
17th stage
Frame blower
Bearing Tunnel
cooling
The net revenue lost due to unplanned
outages is $775,000 per year for the
250 MW F-class plant, or roughly 4-5
percent of net revenue income.** Grace and Christiansen, “Quantifying the cost of
unplanned outage events for combined-cycle
plants,” Energy Tech, August 2012
Source: GE P&W,2013
Slide Courtesy of Peter Evans
Advanced monitoring and analytics
Turbines monitored
~1,550 units globally
24x7x365 coverage
Slide Courtesy of Peter Evans
Are Amazon and Walmart in the Same Business?
20
Social media / webJob search / workE-commerce Tools / cloud / big data
Payments
API Clusters
Messaging services
Source: Rahul Basole and Peter Evans,
with data from ProgrammableWeb,
Center for Global Enterprise, 2015
Enterprise / storage
API Economy Visualized:
API Economy Visualized: Amazon vs Walmart
Walmart
Amazon SNS
Alexa Web Inform
Amazon Marketplac
e
Amazon SimpleDB
Amazon Product
Advertising
Amazon CloudWatc
h
Amazon
Flexible
Amazon Redshift
Amazon SC2
Amazon S3
Amazon Mechanical
TurkAmazon
RDSAmazon DynamoDB
Amazon Queue Service
Social media / web
Job search / work
E-commerce
Tools / cloud / big data
Payments
API Clusters
Messaging services
Walmart
Amazon
Companies
Enterprise / storage
Source: Rahul Basole and Peter Evans,
with data from ProgrammableWeb,
Center for Global Enterprise, 2015
Why has Healthcare been slow to adopt Platforms?
• Conservative “do no harm” culture
• Highly fragmented
• Highly regulated
• Strong incumbents
• Complex data (e.g., compare DHMC to Airbnb)Healthcare’s Transformation into the “Pinnacle” Platform Industry
Vince Kuraitis Jul 22, 2016
Drivers of change
• Shifting Payment — From “Volume to Value”
• Digitization, Interoperability, Mobility
• Democratization of Data
• Obvious value in coordinated care, compliance
• Application of ML/AI
• Blockchain/Crypto
Healthcare’s Transformation into the “Pinnacle” Platform IndustryVince Kuraitis Jul 22, 2016
Challenges with Data
Healthcare’s Transformation into the “Pinnacle” Platform IndustryVince Kuraitis Jul 22, 2016
Questionable uses of consumer data
Privacy concerns coming back• Browsing data
• Education level
• Marital status
• Number of children
in the household
• homeownership
• Mortgage amount
• Vehicle details
• Investment data
• Recent online
purchases
• Interests/hobbies
• Diet purchases
• Charitable donations
• Text messages
• Religion
https://www.nytimes.com/2013/09/01/business/a-data-broker-offers-a-peek-behind-the-curtain.html
GDPR is here
EU GENERAL DATA PROTECTION REGULATION (GDPR)
Expanded scope: GDPR regulation includes “processors” and “controllers” of data.
Significant fines: non-compliance can incur a fine of either 4% annual global revenue or €20 million, whichever is greater.
Increased consumer rights: GDPR has expanded the scope of rights for data subjects, including data portability and access.
Data breach reporting: notification of data breach w/in72 hrs. Data protection officers: institutions may be required to
appoint data protection officers to monitor compliance.
2017: U.S. goes in a different direction4/5/17, 5(00 PMHouse Votes to Let Internet Providers Sell User Browsing Data Without Consent | Nat ional News | US News
Page 1 of 6ht tps:/ /www.usnews.com/news/nat ional- news/ar t icles/2017- 03- 28 /house-…es- to- let- internet- providers- sell- user- browsing- data- without- consent
House Votes to Let Internet ProvidersSell User Browsing Data WithoutConsent
Republicans send bill to President Trump after
dismissing privacy concerns.
President Donald Trump is expected to sign legislat ion that will allow internet providers to sell users' information
without their approval. (Cat Gwynn/Getty)
House Republicans approved legislation Tuesday that would allow internet
service providers such as Comcast and Verizon to sell browsing information
without users' consent.
Data characteristics
• Non-rivalrous use
• Replicated at low marginal cost
• Intermediate good (needs further processing to be useful)
• Difficult to prevent leakage
Protection enforced through legal institutions as well as reputation
Data as a complement
• Network effects: users create value for other users
i.e., network value V(2) > V(1) + V(1)
• Implication: users (individuals or machines) aren’t as valuable alone.
• Implication: users aren’t valuable independent of some platform that can use of data.
Proportion of Tangibles/Intangibles have traded places
Source: Oceantomo
Components of S&P Market Value
Establish property rights in data
• Blockchain as distributed ledger technology might establish data provenance to enable property rights
• Goal: establish verified contractual terms around allowable use and duration of use
– E.g., ThyssenKrupp contracts for IoT data
• Challenge: solve (derivative use) tracking problem
– E.g., summary statistics might suffice; hard to track provenance
A few problems (and more than a few PhD theses)
• Can Blockchain work?– computation, energy
• Are value calculations computationally feasible?
• How to handle user asymmetries?
• Are there directionally correct mechanisms (e.g.,ASCAP) that might be simple to implement and roughly fair?
• Data governance capability must be invested in and managed to solve collective action problems (Ostrom)
• Once in a generation opportunity to reshape business
• Incumbents have strong assets: technology and customer relationships
• Partnerships critical; not every firm will be a platform
The Situation
• Map the Ecosystem
–Where is value being created?
–Where is value being commoditized?
• Where can you participate and control value?
• At each position, decisions and strategy must be set• Build, Partner, or Participate
Manager’s challenge
Strategies to avoid commoditization by larger platforms
• Maintain standalone presence & minimize MSP dependence
• Avoid price competition on MSPs – that’s their game
• Fight for (shared) control over customer relationship
• Narrower scope (specialize), greater depth
• Platform arbitrage
• Build Platforms on top of Platforms
• Buying hardware/infrastructure is the easy part
• Firms struggle to add staff that can leverage data
• Expect organizational pushback when trying to
–Change business models
–Cannibalize existing business
• Risk of doing too little too late
Final considerations
Geoffrey ParkerProfessor, Thayer School Engineering, Dartmouth College
Director, Master of Engineering Management Program
Research Fellow, MIT Initiative on the Digital Economy
@g2parker
Q & A