Vorlesung 6: Digital Platform Governance Flashcards

1
Q

Establishing successful multisided platforms

A

Critical issues that multi-sided platforms must address:
▪ Getting the pricing structure right: critical for the establishment and for long-term profitability
▪ Recognizing the opportunity: when frictions keep market participants from dealing with each other easily and directly transaction costs can be reduced
▪ Securing critical mass: solve the chicken-egg problem
▪ Governance: how participants interact with each other in physical or virtual places
▪ Managing the broader ecosystem: other firms, governments, regulation, and other institutions

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

Platform emergence

A

▪ Platform has to balance need for architecture with randomness and luck of the emergence of their own accord
▪ Users should be provided guidelines but be allowed to take the platform in new directions
▪ Emergence: a small set of micro-level rules gives rise to macro-level movements on emergent systems
▪ Tools-Rules-Interaction-Experience (TRIE) Framework
- Tools and Rules: build out the technology (tools) and
structure the constraints and algorithms that facilitate behavior (rules) by setting boundaries, e.g., video uploading, 280-character limit on Twitter X
- Interaction: structure of value creation, i.e., of the actions that users perform
- Experience: determined by the core value unit on the platform, different use cases

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

Drivers of interactions on platforms

A

Two core principles of platform scale:

Build incentives and behaviors to encourage an ecosystem of producers and
consumers to interact often
▪ Connection: must pre-exist between participants, e.g., Facebook, LinkedIn
▪ Content: no prior relationship necessary for creation and consumption of content, e.g.,
YouTube, eBay
▪ Clout: certain participants have greater influence, e.g., Twitter X (<1% of users have >
10,000 followers)
▪ Coordination: interactions arise from a set of instructed actions, e.g., Wikipedia
▪ Competition: dominated by competitive moves, e.g., 99Designs
▪ Culture and code: centered around shared culture and code, e.g., Reddit

Increase repeatability and efficiency of interactions over time by removing
barriers
▪ Skill barrier, e.g., provision of tools
▪ Time or effort barrier, e.g., efficient aggregation
▪ Investment barrier
▪ Resource barrier, e.g., Amazon Web Services
▪ Access barrier, e.g., Kickstarter

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

Balancing quality and quantity of interactions

A

▪ Platforms must be designed in a manner that optimally balances the quality and quantity of interactions by balancing traction and friction
▪ Design considerations for
friction:
▪ Source of quality
▪ Source of superior signaling
▪ Barrier

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

Interaction ownership

A

▪ Encouraging repeatable interactions isn’t useful unless the platform can own the interaction and prevent off-platform collusion
▪ Connecting buyers and sellers directly before charging the transaction cut weakens the platform’s ability to capture value
▪ A monetization model that involves extracting a cut from the buyer-seller transaction requires a mechanism for owning the end-to-end interaction
▪ Solving the interaction ownership:
- Exchange tracking tools
- Workflow management tools: help to capture repeat interaction, involve a learning curve
and increase multihoming costs
- Reputation as a source of value

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

Suggestions for matchmaking

A

Making markets thick
Screening participants
▪ Account for heterogeneity among participants on both sides, and for changing
participants’ interests and desires
▪ Screen participants by offering something only the desired participants want, or by
signaling (e.g., FarmersOnly.com)
▪ Reasons to limit: limit competition, subsidize marquee users, congestion/ physical
limits
Searching and matching
▪ Make sure interactions happen between participants
▪ Facilitate interactions by organizing participants around a standard that they all agree
to use, e.g., software development kits, 280-character limit of Tweets
▪ Highly complex computerized algorithms for matching, e.g., Google Search
▪ Search buttons
Balancing externalities
▪ Compensate customers, e.g., media businesses create content that they use to
compensate people for receiving advertising messages
▪ Engage in search diversion or recommendations

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

Matchmaking on online-to-offline platforms

A

▪ Online-to-offline (O2O) platforms: offers are not digital (e.g., software, music), but in form of products and services (e.g., cars, apartments)
▪ Increase efficiency and utilization of assets
▪ Handle huge volumes of information:
▪ Personal data, choices and activities, availability and pricing of goods and services,
payments, problems→machine learning
▪ All relevant and useful information can be everywhere on the platforms, all the time
▪ Central task: matching supply and demand over time is difficult because of constraints in real-world, i.e., finite capacity, inventory must be managed
▪ Revenue management improves the more data is available

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

Filters to find matches

A

▪ Platforms solve the abundance problem by creating and managing consumption filters
▪ Units are produced and served to customers based on how well they pass through certain filters, so the results are most relevant to customers
▪ Key factors for passing through filters:
▪ Overlap between what provider provides and customer needs
▪ Data: value units carry data and are matched against the data in a consumer’s filter
▪ Configuration of filters:
▪ Pull versus push, e.g., search queries versus newsfeed
▪ Point-in-time versus cumulative, e.g., based on search history
▪ Active intent versus passive context, e.g., search queries versus newsfeed ▪ Static versus dynamic context, e.g., canned filters
▪ Standalone versus collaborative, e.g., recommendation based on
reviews of similar customers
▪ The network itself is a filter

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

Governance

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

Complementor dedication

A

▪ Key goal of platform governance: complementor dedication
▪ Necessity to balance satisfying global ecosystem needs and local partnership
needs
Global Ecosystem Local Partnership Needs Needs
▪ Adequate rules (as perceived by complementors) increase dedication with
three functions:
- Protection of complementor’s interest vis-à-vis the platform owner
- Prevention of inappropriate behavior on the part of the platform owner
- Assurance of receipt of promised partnership benefits from the platform owner

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

Digital platforms’ boundaries

A

Scope of the platform firm
▪ what assets are owned
▪ what labor is employed
▪ what activities are performed by the firm
Configuration and composition of the platform’s sides
▪ which distinct groups of customers have access to the platform
Digital interfaces
▪ that specify the 2-way exchange of data between the platform firm and each of its
sides

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

Digital platforms’ boundaries
Illustration

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Motivation for platform owner entry

A

Should platform owners still offer complementary products by themselves after the platform has taken off?
Example: Intel versus Amazon
▪ Intel does not want to upset their complementors because they made Intel-specific
investments
▪ Amazon competes on its own platform, retailers did not make Amazon-specific
investments
Motivations for platform owner entries: ▪ Capture value
▪ Spur complementors’ innovations by introducing competition
▪ Exercise better quality control, minimize counterfeiting
Entry decision is related to whether a firm wants to become more like a vertically integrated firm or a multisided platform

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Impact of platform owner entry

A
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Creating cumulative value on platforms (ecosystem)

A

▪ Multihoming costs are decreasing due to social graph (SSO), mobile-based access→erodes lock-in effect of networks
▪ Platforms today need to create cumulative value additionally to network effects to secure customer loyalty
▪ Means:
▪ Reputation
▪ Influence, e.g., followers, access to new rights
▪ Collections
▪ Learning filters, e.g., feeds become more relevant as more information
about user are collected

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Key Business model attributes (Täuscher/Laudien)

A

Value creating dimension:
- Platform Type (web/mobile)
- Key activity (data/community/content)
- price discovery (fixed/seller/buyer/auction/negotiation)
- review system (user/marketplace/none)
Value delivering dimension:
- Transaction content (product/service)
- Transaction type (digital/offline)
- Marketplace participant (c2c/b2c/b2b)
- geographic scope (global/regional/local)
Value capture dimension:
- Key revenue stream (commision/subscription/ads/service sale)
-revenue source (seller/buyer/3rd party/none)

17
Q

How to solve the chicken-egg problem?

A

Platform launch strategies:
(1) Single Target group: focus on a city, use case or industry to reduce the initial market size and therefore required critcal mass
(2) Platform staging: traditional vendor-based business model until critical mass of users is acquired. Switch to platform model after
(3) User Data Portability: reduce entrance barrier by allowing to import data from other platforms (e.g. ratings/reputation)
(4) Subsidizing: subsidize one side of the market to generate critical mass and introduce commision structure

18
Q

When is it optimal for a second-mover to enter the market?

A
  • sufficiently high rate of quality improvement
  • demand window is sufficiently long
  • market growth is slow enough to improve the product quality and catch up to the first-mover
  • the stronger the network effects, the earlier a second competitor needs to enter the market. Otherwise the first-mover can abuse lock-in effects.
19
Q

Platform Boundaries

A

Platform boundaries are
- Scope
- Interfaces
- Sides
-> Differentiate between transaction/innovation platforms in the launch/Maturity phases.

20
Q

Impact of platform owner entry

A
  • positive effect of platform users
  • mixed short-mid term effects on complementors:
    (+) spurs innovation due to competition
    (+) increases platform user base and value capturing opportunities
    (-) discouragement of complementors
    (-) shift of innovation (to avoid competition)
21
Q

Defense strategies of complementors against owner entry

A
  • avoid direct competition by reallocating resources
  • form strategic alliances with platform
  • stop cooperating in the platform (switch, if a viable alternative exists)