Photo by Fahim Muntashir / Unsplash
Who Captures the Surplus? Complementors vs. Platforms — Platform Economics, Bargaining Power & Value Capture
In today’s digital ecosystem, businesses are increasingly enlisting the help of complementors–third-party providers who expand the functionality of a platform. Yet economics tells us that it’s the platform itself that get the majority of the value created, not the complementor. Indeed, this is the essence of Platform Economics, a concept that focuses on platforms' abilities to generate value through mediated interactions and to extract surplus by retaining bargaining power over such interactions.
The Platform Tax: When Fees Absorb Complementor Margins
Complementors need access to large user bases and key infrastructure – be it analytics solutions, ad-tech integrators, or developer utilities that connect into giants like Reddit or Meta. However, the most common way that a platform will charge a platform tax is through:
API access (API usage fee or tiered based usage),
Revenue share deals (e.g. Apple's 15-30% cut),
Commission rates and monetization
For comparison, the premium API of Reddit that year was $0.24 per 1000 calls. One third-party app estimated API costs at almost $20 million per year and eventually shut down. Complementors were discovering that a change in one parameter could destroy business viability overnight.
Similarly, X (formerly Twitter) rolled out steep enterprise-level API tiers-if reports are to be believed, reaching price points of $42,000 per month, which has put many developers and startups in the cold.
These cases emphaseize the power of the platforms to unilaterally increase fees at the expense of complementors, or expelling them altogether, if the complement product does not have its own independent customer base.
Hold-Up Risk and Contracting Safeguards
This asymmetry leads to hold-up risk: opportunistic term changes by platforms once a complementor has done significant, irreversible integration or API-dependent development.
To protect against this, complementors may bargain:
Minimum guarantees of access or price stability;
Most-favored-nation (MFN) by stipulation to be at parity with other providers
Data escrow (or portability) arrangements to preserve historical information if access is withdrawn; and
Service-level agreements (SLAs) with credits or compensation if outages or degraded access occur
But, in practice, complementors - especially if they lack bargaining power - are often forced to accept the platform's terms as they are presented, further increasing their dependency and diminishing their bargaining power in the future.
Value Dependence: Analytics and Ad-Tech Without the Graph
In the advertising and analytics sectors, tools usually have very little value independent of the underlying social graph or ad exchange infrastructure of big platforms.
Ad-tech suppliers are dependent on volume and targeting that can only be delivered through a platform. The whole model falls apart if granular data from users' vanity or behavior is not available.
Analytics integrators rely on consistent APIs and metrics provided by platforms for their operations to work. As the platforms change their data schema or restrict endpoints, downstream value evaporates.
This is an important economic insight: complementors are symbiotic, but the platform owns the lifeline. When the platform pulls back access at the subsistence-level, the complementor's value proposition is lost.
Economic Forces Behind This Dynamic
Why, then, are platforms taking such an enormous share of the value?
Network effects and scalability: Platforms are home to large user bases that, by nature, gives them greater bargaining power. Complementors tend to be more numerous but more fragmented.
Information asymmetry: Platforms have control over usage and pricing analytics. And they can dynamically adjust terms based on third-party performance and dependence.
Switching costs - Complementors invest in workflows, brand integration, and developer tooling that are customized to their clients - all making switching to another platform full of frictions and costly.
These forces are compounding forces working towards a tighter control by platforms while also making it difficult for complementors to exit–even when margins are low.
Real-World Illustrations
Reddit API overhaul (2023): This policy change killed many third-party apps and services, and showed how a single term change can have an immediate and direct impact on the economy of third-party apps and services, and force them out of the ecosystem.
App store regulation under the EU's DMA: Retailers such as Apple are being pressured to reduce commission (from 30% to 15% in some cases). The adjustment is a response to realizing that unregulated platform taxation can distort competition and field monopolization.
Implications and Policy Considerations
For complementors, the takeaways are:
- Don't rely on one channel, diversify your channels and minimize dependency risk.
- Negotiate terms of protection whenever possible even for smaller scale integrations.
- Build proprietary differentiation (data, analytics, UX) that will survive if access to the platform is throttled.
For policymakers and platform architects, the goal is to maintain contestability, so that complementors have viable options to choose from. This also means there is less lock-in through portability standards, so complementors can negotiate more equitably.
Also indicative is that all developer onboarding documents begin with instructions on how to get reddit api key - a reminder that access is the scarce valuable resource over which the platform holds bargaining power.
Conclusion
Platform economics reminds us that value creation does not guarantee equal capture. When complementors depend on a platform’s infrastructure, the platform is able to gain disproportionate surplus–usually through opaque or shifting fee schedules. Understanding the dynamics of bargaining power and hold-up risk and dependency are important for complementors as they strive to develop contractually robust business models in the platform economy.