Platform Lock-In and Switching Costs in Creator Economies: Measuring Audience Mobility, Bargaining Power, and the Limits of Data Portability

Abstract

Creators on YouTube, Substack, TikTok, and adjacent platforms often appear to possess unusually high autonomy: they choose topics, set schedules, and monetize direct relationships with audiences. Yet that autonomy is conditional. Once a creator has accumulated followers, reputation, monetization routines, and platform-specific archives, the practical difficulty of moving elsewhere can be substantial. This article develops and demonstrates a measurement framework for analyzing platform lock-in in creator economies. Building on switching-cost theory, platform economics, and digital labor scholarship, the study introduces the Creator Switching Cost Index (CSCI) , which decomposes platform dependence into five dimensions: technical migration, audience reconstruction, monetization reconfiguration, algorithmic/reputational loss, and behavioral-habitual costs. The analysis is based on a structured comparative reading of platform affordances and creator workflows across YouTube, Substack, and TikTok, alongside the established literature on two-sided markets, network effects, and platformization. The results suggest that the principal barrier to exit is rarely the raw transfer of files. Rather, the binding constraint is the loss of audience continuity, discovery history, and revenue relationships that are embedded in platform infrastructures. YouTube and TikTok exhibit especially strong lock-in because audience access is mediated by recommendation systems and engagement histories that do not travel with the creator. Substack reduces some technical switching costs through email-based distribution, but it still creates meaningful lock-in through subscriptions, payment rails, and platform-mediated discovery. The article argues that data portability rules are necessary but insufficient: they can lower transfer frictions, yet they do not fully move social capital, algorithmic reputation, or monetization relationships. The policy implication is that effective governance should combine portability with interoperability, payment portability, and greater transparency about ranking and monetization systems.

Keywords: platform economics; switching costs; creator economy; digital labor; market power; data portability

Introduction

Creator economies and the platformization of cultural labor

The contemporary creator economy is built on an apparent paradox. On the one hand, platforms such as YouTube, Substack, and TikTok promise low barriers to entry, direct audience access, and flexible monetization. On the other hand, the very same platforms reorganize cultural production into highly dependent, infrastructure-intensive forms of labor. The creator is not simply a producer of content; the creator is also a manager of attention, a curator of data, a buyer of platform visibility, and, often, a self-exploiting labor force that must remain continuously legible to algorithms and audiences alike (Cunningham & Craig, 2019; Duffy, 2017; Nieborg & Poell, 2018; van Dijck et al., 2018). This dual role complicates familiar assumptions about entrepreneurship and autonomy. Creators can appear independent while remaining structurally dependent on a platform's discovery system, moderation regime, monetization rules, and analytics interface.

This dependency is best understood through the lens of platformization , the process by which digital intermediaries become increasingly central to the organization of cultural production, circulation, and monetization (Poell et al., 2019). Platformization does not merely provide a distribution channel; it embeds creators within a governance architecture that shapes what can be seen, shared, monetized, and retained. Gillespie (2018) argues that platforms are custodians of public speech, while Zuboff (2019) shows how platform business models extract behavioral surplus from user activity. For creators, the consequences are especially pronounced because their work is already tied to metrics of visibility, engagement, and audience growth. The creator economy therefore occupies an important analytical position between digital labor and platform economics: it is a labor market organized through privately controlled infrastructures, but it also resembles a market for attention assets that can be accumulated, degraded, or trapped.

Within this environment, the central question is not merely how creators gain audiences, but how easily those audiences can be moved. A creator who has spent years building a subscriber base, a comment community, or a subscriber-funded newsletter may find that leaving a platform means losing more than content. The creator can lose discovery history, audience routines, payment infrastructure, search visibility, reputation signals, and monetization contracts. These losses are not incidental. They are the practical expression of switching costs —the costs borne by users when moving from one supplier or platform to another (Klemperer, 1987; Farrell & Klemperer, 2007). In the classic industrial-organization literature, switching costs generate lock-in, dampen competition, and can confer market power on incumbents. In creator economies, the stakes are more complex because the “product” that might switch is not merely a consumer account; it is also a labor process and a social graph.

Platform lock-in as a creator-specific switching problem

Much of the switching-cost literature was developed in contexts where consumers choose among standardized goods or services, such as banking, telecommunications, or software. In those settings, the problem is relatively straightforward: consumers face search costs, learning costs, contractual penalties, and data-transfer frictions. Platform economies, by contrast, operate in two-sided or multi-sided markets where one side's participation depends on the participation of the other side (Armstrong, 2006; Rochet & Tirole, 2003). Creators occupy an especially interesting position in these markets because they are both labor providers and demand attractors. Their audience creates value for advertisers, subscribers, and the platform itself, while the platform in turn controls the means by which that audience is reached.

This makes creator switching costs more than a simple file-transfer problem. The creator's value is often deeply relational. A subscriber list is valuable because subscribers trust the creator and expect future communication. A YouTube channel is valuable because it has accumulated watch history, recommendations, and a back catalog that feeds discovery. A TikTok presence is valuable because the platform's algorithm has learned a creator's engagement profile and can distribute content accordingly. When a creator exits, these assets are not fully portable. They may be partially exportable as metadata, but their behavioral and algorithmic value is usually platform-specific. Thus the real switching cost is not just the loss of content but the loss of the feedback loop that made the content discoverable in the first place.

From a market-power perspective, this distinction matters. If creators cannot credibly threaten to leave, platforms may be able to reduce revenue shares, alter ranking rules, tighten moderation, or shift monetization policies with limited risk of supply loss. Classic switching-cost theory predicts exactly this kind of asymmetry: the side with weaker outside options is less able to bargain over price and quality (Farrell & Klemperer, 2007; Klemperer, 1987). In creator economies, the outside option is often underspecified because creators do not own the infrastructure through which audiences are encountered. The platform can therefore extract rents not only through explicit fees, but also through opaque changes in visibility and monetization conditions. Stigler (1961) anticipated this dynamic in his analysis of information asymmetry: where consumers or producers cannot cheaply observe quality or price changes, market discipline weakens.

Why YouTube, Substack, and TikTok provide a useful comparison

YouTube, Substack, and TikTok represent three distinct but overlapping creator infrastructures. YouTube is a mature video platform in which creators benefit from a large archive, searchability, and multiple revenue channels, including advertising, memberships, sponsorships, and direct support. Substack is built around email newsletters and paid subscriptions, making it seem closer to a direct-to-audience publishing system than a conventional social network. TikTok is a short-form video platform whose most powerful distribution mechanism is algorithmic recommendation, not follower-based subscription. These differences make the trio analytically useful. YouTube emphasizes archival depth and algorithmic discovery; Substack emphasizes direct subscription and email portability; TikTok emphasizes short-form virality and recommendation intensity.

These are not merely technical distinctions. They generate different kinds of switching costs. A YouTube creator may have a large archive but still remain dependent on recommendation systems for reach. A Substack writer may control an email list but still rely on the platform for payment handling and discoverability. A TikTok creator may acquire visibility rapidly but remain heavily dependent on the platform's ranking logic, trend cycles, and format conventions. The comparison therefore helps identify which parts of creator dependence are due to the business model, which are due to technical architecture, and which are due to audience behavior. It also clarifies a key policy question: if data portability rules are designed primarily around file export, do they address the costs that matter most?

Research questions and contribution

This study addresses three research questions:

• RQ1: Which platform features generate the largest switching costs for creators?
• RQ2: How do these switching costs differ across YouTube, Substack, and TikTok?
• RQ3: Which policy instruments are most likely to improve creator bargaining power by reducing lock-in?

The article makes three contributions. First, it develops a multi-dimensional switching-cost framework tailored to creator economies rather than generic consumer markets. Second, it offers a structured comparative assessment of three platform archetypes that often appear similar in policy debates but differ substantially in the distribution of lock-in. Third, it evaluates data portability as a governance response and argues that portability alone is insufficient when the most valuable assets are relational, algorithmic, and monetization-specific rather than merely informational.

Figure 1 summarizes the conceptual logic that motivates the analysis.

Methodology

Research design

The study uses a structured comparative design to develop and demonstrate a measurement framework for creator switching costs. Rather than estimating a population average from proprietary platform telemetry, the analysis draws on a triangulated evidentiary base: peer-reviewed literature on switching costs, two-sided markets, platformization, and digital labor; public descriptions of platform business models and creator workflows; and a theory-driven comparison of platform affordances. This approach is appropriate because the central analytical object—switching costs in creator economies—depends heavily on the interaction between platform architecture and user behavior, both of which are observable in public-facing features even when internal data are unavailable.

In methodological terms, the article is best understood as an original measurement study and conceptual demonstration. It does not claim to observe the average creator's actual exit behavior across the entire creator economy. Instead, it produces a reproducible framework that can later be implemented using survey data, platform traces, or administrative records. The absence of those data in this article is not a weakness so much as an important reminder of the governance problem itself: creators commonly face the same opacity when deciding whether to leave a platform.

Data sources and evidentiary base

The comparative analysis relies on four types of evidence. First, it uses established economic theory on switching costs, network effects, and market power to define the underlying mechanisms (Farrell & Klemperer, 2007; Klemperer, 1987; Rochet & Tirole, 2003). Second, it incorporates platform theory and platformization scholarship to explain how digital infrastructures shape cultural production (Jacobides et al., 2018; Nieborg & Poell, 2018; Poell et al., 2019; van Dijck et al., 2018). Third, it draws on scholarship on digital labor and creator work to identify the specific forms of dependence experienced by creators (Abidin, 2018; Cunningham & Craig, 2019; Duffy, 2017; Terranova, 2000). Fourth, it uses policy and governance literature to evaluate portability and interoperability proposals (Crémer et al., 2019; Stigler Committee on Digital Platforms, 2019).

The article treats YouTube, Substack, and TikTok as platform archetypes. YouTube represents archive-intensive, recommendation-mediated video production. Substack represents subscription-based newsletter publishing with a relatively open distribution standard. TikTok represents recommendation-intensive short-form video production where visibility is deeply contingent on the platform's ranking architecture. The point is not that every creator experiences these platforms identically. Rather, the platforms differ in ways that are analytically useful for understanding distinct forms of lock-in.

Operationalizing switching costs

To make the concept measurable, the study proposes the Creator Switching Cost Index (CSCI) . The CSCI decomposes switching costs into five dimensions that are especially salient in creator economies:

• Technical migration cost: the effort required to move content, metadata, archives, and workflows to another service.
• Audience reconstruction cost: the loss of followers, subscribers, referral traffic, and audience routines when moving off-platform.
• Monetization reconfiguration cost: the effort required to rebuild revenue streams, payment rails, sponsorship routines, and subscription relationships.
• Algorithmic and reputational cost: the loss of ranking history, engagement signals, moderation status, and discoverability.
• Behavioral and habitual cost: the friction created by re-learning production routines, audience expectations, and content norms.

Each component can be coded on a normalized scale from 0 to 1, where 0 represents minimal cost and 1 represents maximal cost. The overall index is defined as follows:

(1)

In Eq. (1), is the value of cost dimension for platform or creator case , and is the weight assigned to that dimension. The baseline specification uses equal weights, which is a conservative assumption because it gives technical transfer costs the same importance as audience and monetization costs. In practice, many creators are likely to experience the latter as more binding. For that reason, the study also considers two alternative weighting logics in sensitivity analysis: a discovery-heavy scheme, in which algorithmic and audience costs receive greater weight, and a monetization-heavy scheme, in which subscription and payment dependence are emphasized.

Table 1 summarizes the dimensions and the logic of the index.

Outside options and creator bargaining power

To connect switching costs to bargaining power, the study adapts the outside-option logic common in industrial organization. The intuition is simple: when the cost of leaving a platform rises, the value of the creator's alternative options falls. This can be represented as:

(2)

In Eq. (2), is the bargaining power of creator or platform case , and is the value of the creator's outside option. As the CSCI increases, bargaining power declines, all else equal. This expression is intentionally stylized. It does not claim that bargaining power is reducible to a single ratio, only that platform lock-in and outside options move in opposite directions. The equation is useful because it makes clear that a creator with a strong, portable audience asset—such as a widely subscribed email list or a direct-commerce channel—has a stronger negotiating position than a creator whose audience is almost entirely platform-native.

A related way to interpret the same logic is through expected revenue retention after switching:

(3)

In Eq. (3), is pre-switch revenue, is expected post-switch revenue, and captures the extent to which audience stickiness magnifies the effect of switching costs. The equation formalizes a common empirical intuition in creator work: even when content can be re-uploaded, revenue does not necessarily follow. Reach, trust, and monetization can decay faster than the archive itself.

Analytic strategy and sensitivity checks

The analysis proceeds in three steps. First, it identifies platform features that affect each switching-cost dimension. Second, it compares the platforms qualitatively and orders them by relative lock-in burden. Third, it evaluates whether policy responses target the components that matter most. Because the article is theory-driven rather than estimate-driven, sensitivity analysis is conducted conceptually by varying the relative importance of the switching-cost dimensions. The principal substantive question is whether the ranking of platforms changes when audience and monetization costs are weighted more heavily than technical migration costs. The answer, discussed below, is that the broad ordering is robust: the most important costs are not the easiest ones to port.

Results

Switching costs are dominated by audience and monetization loss, not by file transfer

The most important result is that creator switching costs are not primarily about moving files. The raw content archive matters, but it is rarely the binding constraint. Instead, the central loss is audience reconstruction : the creator must persuade people to reassemble elsewhere, at the moment those people are most likely to remain where they are. This aligns with the classic switching-cost literature, which shows that firms often retain customers not because moving is technically impossible, but because leaving entails forfeiting familiarity, accumulated benefits, and relationship-specific capital (Farrell & Klemperer, 2007; Klemperer, 1987). In creator economies, the “benefit” is often not a discount or loyalty point but an audience relationship and a stream of recurring attention.

A second result is that monetization reconfiguration is almost as important as audience reconstruction. Creators do not simply want to move their content; they need to move their revenue model. Sponsorships, subscriptions, affiliate arrangements, live-gifting routines, memberships, and ad monetization all depend on platform-specific rules. A creator may be able to transport a video or newsletter issue, but that does not guarantee revenue continuity. This is a crucial reason why market power in creator economies is often understated: the platform does not need to own the creator's ideas if it can control the infrastructure through which those ideas are monetized.

A third result is that algorithmic and reputational costs create a hidden layer of dependence. Search rank, recommendation history, and engagement signals accumulate over time and cannot be simply exported. This is especially visible in platforms where discovery is algorithmically mediated rather than follower-based. Cotter (2019) shows that influencers' visibility is often negotiated through opaque algorithmic systems, and Gillespie (2018) demonstrates that content governance itself is a form of power. For creators, the practical consequence is that audience visibility is often a platform-owned asset, not a transferable one.

Comparative platform profiles: YouTube, Substack, and TikTok

Table 2 presents the comparative assessment of the three focal platforms. The table uses qualitative categories rather than pseudo-precise numeric scores because the article's purpose is to expose the structure of lock-in, not to fabricate artificial measurement certainty. Even so, the pattern is clear. YouTube and TikTok are high-lock-in environments, albeit for different reasons, while Substack is less technically sticky but still far from frictionless.

Note: Categories are author-generated and based on a comparative synthesis of platform architecture, creator workflows, and the literature on platformization and switching costs.

For YouTube, the main source of lock-in is archival accumulation combined with algorithmic discovery. The creator's back catalog, comment history, and recommendation profile all deepen over time, which means that exit entails not only a content re-upload but also a partial loss of historical visibility. Burgess and Green (2009) show how YouTube combines participatory culture with platform control; the present analysis extends that insight by emphasizing how the archive itself becomes a barrier to mobility. A YouTube creator can upload elsewhere, but the past that made the channel legible to viewers does not transfer in full. Because discovery is partly linked to watch history, channel authority, and recommendation momentum, the creator's labor is entangled with platform memory.

Substack behaves differently. Its email-based architecture makes it easier to claim a direct relationship with readers, and email is a comparatively open distribution standard. That reduces technical lock-in and, in principle, lowers audience reconstruction costs. Yet Substack still creates meaningful dependence through subscriptions, payment processing, platform recommendations, and the accumulated trust embedded in recurring newsletters. The writer may own the list, but the economics of the list remain platform-mediated. In practice, a move away from Substack can mean rebuilding payment relationships, re-establishing brand continuity, and persuading readers to adopt a new routine. The platform therefore sits in a middle position: less sticky than a fully closed social feed, but not nearly as portable as advocates sometimes suggest.

TikTok exhibits the strongest lock-in in the comparison. Short-form video is highly portable as a file format, but the value of a TikTok presence is not the file. It is the interaction between trend timing, platform-native editing norms, audience expectations, and algorithmic distribution. Creators can re-upload clips elsewhere, but they cannot export the For You Page. This matters because the platform's ranking system is itself a source of accumulated advantage. As on other platforms, creator visibility depends on engagement histories that are not fully transferable (Cotter, 2019; Gillespie, 2018). The result is a highly asymmetric exit problem: creators can leave TikTok, but the audience-growth mechanism that made them viable there does not travel cleanly with them.

Robustness of the platform ranking under alternative weightings

The qualitative ranking is robust to plausible alternative weightings. When technical migration is weighted more heavily, Substack appears even less sticky than the other two platforms. When discovery and audience reconstruction are weighted more heavily, TikTok remains the most locked-in environment, followed closely by YouTube. When monetization is weighted most strongly, Substack rises somewhat because subscription revenue creates a relationship-specific dependency that is not reducible to file transfer. Even under this monetization-heavy scenario, however, Substack still does not resemble the high-lock-in profile of TikTok, where discovery and audience formation are tightly coupled to the platform's own recommendation engine.

This robustness matters because it suggests that the key empirical intuition does not depend on a narrow choice of weights. The platform ranking changes only modestly when the analyst shifts from a content-centric to a revenue-centric view. That stability implies that the lock-in problem is genuinely multi-dimensional. It is not enough to ask whether a creator can export content. One must ask whether the creator can export the conditions under which the content becomes valuable . In the creator economy, that is often the decisive question.

Bargaining power is highest when audience assets are portable and diversified

Equation (2) implies that bargaining power rises when outside options are strong and falls when switching costs are high. The comparative analysis therefore suggests a hierarchy of creator bargaining positions. Creators with diversified revenue streams, substantial owned audiences, and multi-homed distribution strategies have the strongest negotiating position. Creators whose income is overwhelmingly dependent on one platform's recommendation system have the weakest. This is consistent with platform economics: the more indispensable the platform becomes to the creator's revenue model, the more leverage the platform obtains over rules, fees, and governance changes (Rochet & Tirole, 2003; Cennamo & Santalo, 2013).

From a labor perspective, this means that “independence” in the creator economy is often conditional on owning some portion of the audience relationship outside the platform. Email lists, direct subscriptions, personal websites, community servers, and payment rails can strengthen outside options. Yet these assets come at a cost: creators must spend time maintaining parallel infrastructures, cross-posting content, and persuading audiences to move between spaces. The same multi-homing strategy that improves bargaining power can also intensify labor, which echoes broader scholarship on digital labor and aspirational work (Duffy, 2017; Terranova, 2000; Wood et al., 2019). In other words, reducing platform dependence often means increasing unpaid infrastructural work.

The results also support a more nuanced interpretation of platform power. Platform control is not limited to overt fee changes. It includes the ability to set the conditions under which creators can become visible at all. In this sense, recommendation systems and moderation systems are not merely technical features; they are bargaining institutions. Gillespie (2018) and Zuboff (2019) both emphasize that digital platforms govern visibility and data extraction. The present analysis extends that point by showing that such governance is particularly consequential for creators because it affects not just speech, but income. A creator who cannot preserve audience attention after switching is negotiating with an outside option that is structurally weak.

Policy scenario analysis: what data portability can and cannot solve

The policy implication that follows is clear: data portability is helpful, but its scope is limited. In theory, portability rules lower technical migration costs by allowing users to export their data. In practice, the most valuable creator assets are not just data objects but relationship structures. A follower list without active follow-through is not an audience. A video archive without recommendation momentum is not equivalent to an active channel. A newsletter subscriber file without ongoing trust is not the same as a functioning membership base. For this reason, portability rules can reduce one layer of lock-in while leaving others largely untouched.

Policy reports on digital platforms have reached similar broad conclusions. The Stigler Committee on Digital Platforms (2019) argues that data portability and interoperability are central to competition policy in digital markets, while Crémer et al. (2019) emphasize the importance of access, interoperability, and business-model constraints in the digital era. The present analysis is consistent with those findings, but it adds a creator-specific twist. In creator economies, the most critical portable asset is not raw content, but the ability to preserve audience continuity, payment continuity, and discoverability continuity. This suggests that a mere download button is insufficient as a remedy for lock-in.

Table 3 summarizes several policy tools and the switching-cost dimensions they are most likely to affect.

Note: The table is an author-generated policy synthesis informed by the switching-cost and platform-governance literature.

Discussion

Platform lock-in is a market-power problem, not just a usability problem

The most important interpretive point is that platform lock-in in creator economies is not simply an inconvenience or design flaw. It is a market-power problem. Once creators have accumulated audience capital within a platform, the platform can shape rents, access, and visibility in ways that would be more difficult in a more contestable environment. The switching-cost literature has long shown that lock-in can dampen competition by weakening the disciplining force of exit (Farrell & Klemperer, 2007; Klemperer, 1987). What is distinctive in creator economies is that the locked-in “customer” is also a worker and, in many cases, a small business. This creates a hybrid form of dependence that sits at the intersection of labor relations and platform governance.

The results also show why classic consumer analogies are incomplete. A consumer who changes streaming services may lose a playlist, but a creator who changes platforms may lose an audience infrastructure. The creator's “asset” is dynamic, relational, and co-produced by the platform's recommendation and monetization systems. This is why platform envelopment is so powerful: firms can expand from hosting content to controlling adjacent functions—payments, analytics, ads, messaging, and subscriptions—thereby increasing the number of points at which switching becomes costly (Eisenmann et al., 2011). The expansion of platform ecosystems is therefore not just diversification; it is a form of dependence deepening (Gawer & Cusumano, 2014; Jacobides et al., 2018).

Why data portability is necessary but insufficient

Data portability is often treated as the obvious policy fix for lock-in. The analysis here suggests a more cautious view. Portability can reduce some barriers, especially technical migration costs, but it does not automatically recreate the mechanisms that made a creator successful on the original platform. Three reasons stand out. First, the value of audience data depends on context; a subscriber exported from one platform may not behave the same way on another. Second, algorithmic reputation is not fully portable; the new platform must still learn the creator's value, and the old platform's accumulated visibility does not transfer. Third, monetization relationships are embedded in specific payment systems, contractual expectations, and audience habits. The result is that portability may lower the cost of leaving without guaranteeing that the creator can thrive elsewhere.

This is particularly visible in Substack-like environments. Because email is an open standard, it appears more portable than a recommendation-led social feed. Yet a newsletter is only partly an email list. It is also cadence, brand, subscription habit, and trust. The same can be said of YouTube and TikTok, though in different ways. For YouTube, the archive and recommendation history are the core of the business model. For TikTok, recommendation intensity and trend alignment are central. In both cases, the creator cannot fully export the platform-specific conditions under which visibility is produced. Data portability addresses a layer of information, but not the social and algorithmic machinery that transforms information into reach.

For this reason, the policy bundle should be broader than portability alone. Interoperability is especially important because it reduces the need for a full exit in the first place. If creators can publish once and distribute across services, the platform's hold

Conclusion

This article has argued that platform lock-in in creator economies is best understood as a multi-dimensional switching-cost problem rather than a simple matter of exporting files or content archives. The central finding is that the most consequential barriers to exit are not technical alone. Instead, they arise from the difficulty of reconstructing audiences, preserving monetization relationships, and retaining algorithmic and reputational capital after leaving a platform. In this respect, the proposed Creator Switching Cost Index (CSCI) captures a form of dependence that is invisible if one looks only at content portability.

The comparative analysis of YouTube, Substack, and TikTok further shows that different platform architectures produce different lock-in profiles. YouTube combines archival accumulation with recommendation dependence, making the back catalog valuable but still platform-bound. Substack reduces some technical friction through email-based distribution, yet it still ties creators to platform-mediated subscriptions, payments, and discovery. TikTok exhibits the strongest lock-in because audience formation is deeply entangled with recommendation systems, trend timing, and short-form engagement logics that do not travel with the creator. Across all three cases, the binding constraint is less the movement of content than the transfer of the conditions that make content visible and monetizable.

These patterns have direct implications for creator bargaining power. Creators who can preserve an outside option—through owned email lists, direct payment channels, multi-homing strategies, or diversified distribution—are better positioned to negotiate changes in fees, moderation, or monetization rules. By contrast, creators whose audiences are concentrated within one platform face weaker bargaining positions because leaving would mean sacrificing accumulated attention and revenue. The result is a distinctive form of labor-market asymmetry: creators may appear entrepreneurial, but their leverage is often constrained by infrastructures they do not control.

From a governance perspective, the findings suggest that data portability is necessary but insufficient. Portability rules can lower technical migration costs, but they do not move algorithmic reputation, audience habit, or platform-specific monetization structures. More effective policy responses would combine portability with interoperability, payment portability, transparency in ranking and monetization systems, and safeguards against retaliatory downranking or discriminatory treatment of multi-homing creators. Future research should move beyond descriptive comparisons and estimate the CSCI using creator surveys, platform trace data, or quasi-experimental policy changes. Such work would help determine not only how lock-in operates, but also which interventions most effectively restore competitive pressure and creator autonomy.

References