Platform ratings and deactivation fairness
1. Introduction
Platform ratings and deactivation fairness concern the rules and procedures used by digital platforms to evaluate workers, sellers, drivers, delivery partners, and other service providers, and to decide whether they should continue using the platform.
Digital platforms such as ride-hailing applications, food-delivery services, online marketplaces, and freelance websites often rely on ratings, customer complaints, cancellation rates, acceptance rates, and automated algorithms to monitor performance.
For example, a driver may receive a low customer rating, a delivery partner may be accused of misconduct, or an online seller may receive negative reviews. The platform may respond by reducing access to assignments, suspending the account, or permanently deactivating it.
Platform ratings and deactivation fairness means that these decisions should be transparent, evidence-based, proportionate, and consistent with applicable law and contractual obligations. Workers and sellers should have a meaningful opportunity to understand and challenge serious adverse decisions.
The issue is particularly important in the gig economy, where deactivation can result in the immediate loss of income and access to customers.
2. Meaning of Platform Ratings
Platform ratings are numerical scores, reviews, or performance indicators used to measure the quality and reliability of services provided through a digital platform.
Common examples include:
Customer satisfaction ratings.
Driver or delivery-partner ratings.
Seller ratings and product reviews.
Cancellation and acceptance rates.
Punctuality and completion rates.
Complaint and refund statistics.
Algorithmic risk or fraud scores.
Platforms may use these indicators to rank participants, distribute work, restrict opportunities, or determine eligibility to remain on the platform.
However, a rating is not necessarily an accurate measure of an individual's performance. Customers may submit unfair reviews, discriminatory complaints, retaliatory ratings, or complaints arising from circumstances outside the worker's control.
A fair system must therefore distinguish genuine performance problems from inaccurate, biased, or manipulated feedback.
3. Meaning of Deactivation
Deactivation is the restriction or termination of a user's ability to access a platform or perform work through it.
It may take several forms:
Temporary suspension: Access is restricted while a complaint or investigation is pending.
Performance-based deactivation: Access is removed because ratings or performance indicators fall below a threshold.
Misconduct-based deactivation: The platform responds to alleged harassment, fraud, violence, theft, or other prohibited conduct.
Automated deactivation: An algorithm makes or triggers the decision based on platform data.
Permanent deactivation: The platform terminates access indefinitely or ends the contractual relationship.
The seriousness of the decision matters. A temporary restriction during a credible safety investigation may require a different procedure from permanent removal based on several low customer ratings.
4. Legal Principles Governing Rating and Deactivation Fairness
A. Transparency
Platforms should explain how ratings are calculated, what conduct may lead to deactivation, and which thresholds or rules apply. Where algorithms are used, workers should receive meaningful information about the factors that materially affect their access to work, subject to legitimate confidentiality and security requirements.
B. Accuracy and reliability
A platform should not rely blindly on a single negative rating, an unverified complaint, or inaccurate data when making a serious decision. Relevant evidence may include trip records, delivery records, customer communications, location information, and the worker's explanation.
C. Proportionality
The response should correspond to the seriousness of the conduct. A minor service complaint may justify coaching or a warning, whereas credible evidence of a serious safety threat may justify immediate temporary suspension.
D. Notice and opportunity to respond
Where practicable and consistent with urgent safety requirements, the platform should notify the worker of the allegation and provide an opportunity to respond before making a final adverse decision.
E. Human review and appeal
A meaningful appeal should allow the worker to challenge incorrect facts, explain unusual circumstances, and request reconsideration by someone capable of changing the original decision. An automated response that merely repeats the initial outcome may be inadequate as a matter of good practice and may not satisfy applicable legal obligations.
F. Non-discrimination
Ratings and performance systems should be examined for discriminatory effects. For example, a driver should not be penalized merely because a customer gives a lower rating due to the driver's disability, race, gender, accent, or another protected characteristic.
G. Consistency
Similar cases should generally be treated similarly. A platform should explain material differences in treatment and should not selectively enforce its rules against particular workers.
5. Detailed Case Laws
The following cases establish important principles relating to platform control, gig-worker status, access to legal remedies, and procedural fairness. Some concern deactivation or dismissal directly; others provide the wider legal framework rather than deciding the fairness of a ratings-based deactivation itself.
Case 1: Uber BV v Aslam [2021] UKSC 5
Court: Supreme Court of the United Kingdom
Facts: Uber drivers argued that they qualified as workers under UK employment legislation and were entitled to statutory protections, including minimum-wage and working-time rights. Uber maintained that the drivers were independent contractors.
Legal issue: Whether the drivers were workers under the applicable legislation, considering the actual relationship between Uber and its drivers.
Judgment: The Supreme Court upheld the finding that the drivers were workers for the relevant statutory purposes. It emphasized the importance of the practical relationship and the degree of control exercised by Uber, rather than treating the written contractual description as decisive.
Relevance to platform ratings and deactivation fairness: The decision demonstrates that a platform's control over access to work, pricing, performance standards, and other working conditions can be legally significant. Where a platform exercises substantial control over workers, the applicable employment protections may be broader than the platform's contractual labels suggest.
The judgment did not establish a universal right to a hearing before every deactivation. Its importance lies in assessing the real nature of platform control and the statutory rights that may follow.
Case 2: Klooger v Foodora Australia Pty Ltd [2018] FWC 6836
Court/Tribunal: Fair Work Commission, Australia
Facts: Franco Klooger worked as a food-delivery rider for Foodora. The dispute concerned his employment status and the termination of his engagement after conflict surrounding his work and activities connected with worker representation.
Legal issue: Whether Klooger was an employee for the purposes of the relevant unfair-dismissal protections and whether the termination could be reviewed under those protections.
Decision: The Commission found that the relationship was employment in substance and that Klooger's termination was unfair in the circumstances. The case is widely discussed in relation to employment status and protections for platform-based delivery workers.
Relevance to platform ratings and deactivation fairness: A platform cannot necessarily avoid employment obligations simply by describing a delivery worker as an independent contractor. If the relationship satisfies the relevant legal test for employment, the platform may be subject to statutory dismissal protections.
This case is especially relevant where a purported performance-based deactivation may actually involve an employment termination or retaliation for worker representation.
Case 3: Uber Technologies Inc. v. Heller, 2020 SCC 16
Court: Supreme Court of Canada
Facts: David Heller, a delivery worker using Uber's platform, challenged an arbitration clause that required disputes to be resolved through an expensive process abroad. The contractual arrangement made pursuing a claim practically inaccessible to him.
Legal issue: Whether the arbitration clause was unconscionable and unenforceable.
Judgment: The Supreme Court of Canada held that the arbitration clause was unconscionable and unenforceable under the applicable law. The cost and practical barriers to arbitration were central to the decision.
Relevance to platform ratings and deactivation fairness: A platform should not assume that a contractual dispute-resolution clause will prevent workers from challenging adverse decisions. Where a deactivated worker has no realistic way to pursue an appeal or legal remedy, the fairness and enforceability of the dispute-resolution arrangement may be questioned.
The case does not establish a particular deactivation procedure, but it highlights the importance of accessible remedies.
Case 4: Independent Workers Union of Great Britain v Central Arbitration Committee [2023] UKSC 43
Court: Supreme Court of the United Kingdom
Facts: The Independent Workers Union of Great Britain sought collective bargaining recognition for Deliveroo riders. The dispute turned on whether the riders were workers within the relevant statutory definition for collective bargaining purposes.
Legal issue: Whether Deliveroo riders qualified as workers entitled to the requested collective bargaining process under the statutory framework.
Judgment: The Supreme Court dismissed the appeal and upheld the conclusion that the riders did not qualify as workers under the relevant statutory provisions, particularly given the genuine contractual right of substitution.
Relevance to platform ratings and deactivation fairness: The case demonstrates that employment and collective rights depend on the actual legal and working arrangements. The existence of platform control alone does not guarantee that every rider qualifies for every statutory protection.
A fair deactivation policy should therefore be assessed alongside the worker's legal status and the protections available under the relevant jurisdiction. This case did not directly decide the fairness of individual ratings or deactivations.
Case 5: Autoclenz Ltd v Belcher [2011] UKSC 41
Court: Supreme Court of the United Kingdom
Facts: Workers engaged in car-valeting services were described in written contracts as self-employed and were subject to provisions that purported to deny employment status.
Legal issue: Whether the written contractual terms accurately reflected the true working relationship.
Judgment: The Supreme Court held that courts and tribunals must examine the true agreement and the reality of the relationship, rather than automatically accepting contractual wording that does not reflect the practical arrangement.
Relevance to platform ratings and deactivation fairness: A platform's terms of service cannot necessarily settle the legal status of a worker. If the actual relationship indicates employment, applicable dismissal, discrimination, or other statutory protections may apply despite contractual wording.
This principle is important when a platform describes account removal as merely terminating access to a service, although the practical effect may be the loss of employment or work-related rights.
Case 6: Pimlico Plumbers Ltd v Smith [2018] UKSC 29
Court: Supreme Court of the United Kingdom
Facts: Gary Smith worked as a plumber under an arrangement that described him as self-employed. The company imposed various working requirements, while the contractual arrangement also included a limited right of substitution.
Legal issue: Whether Smith qualified as a worker under the relevant UK employment legislation.
Judgment: The Supreme Court concluded that Smith was a worker for the relevant statutory purposes, although he was not an employee under the ordinary employment-contract test.
Relevance to platform ratings and deactivation fairness: The case reinforces the importance of examining control, personal service, and the practical nature of a working relationship. Depending on the applicable law and status, a person whose platform account is deactivated may have statutory rights that cannot simply be excluded by contractual labels.
The decision did not concern algorithmic ratings directly, but it is relevant to determining which legal protections may govern platform work.
Case 7: Asociación Profesional Elite Taxi v Uber Systems Spain SL, Case C-434/15 (2017)
Court: Court of Justice of the European Union
Facts: A Spanish taxi association challenged Uber's business model, raising questions about the legal classification of Uber's services.
Legal issue: Whether Uber's service was an information-society service or part of an overall service in the field of transport.
Judgment: The Court held that the Uber service at issue fell within the field of transport services rather than being classified solely as an information-society service.
Relevance to platform ratings and deactivation fairness: The case illustrates that digital platforms can be subject to the regulatory framework applicable to the underlying activity, not merely to rules governing software or online intermediation. The appropriate legal framework for a deactivation dispute may therefore depend on the type of platform, service, and jurisdiction.
The case did not establish a specific right to challenge customer ratings or account deactivation.
6. Legal Position in India
In India, platform ratings and deactivation disputes must be assessed under the laws applicable to the particular relationship. Relevant considerations may include contractual terms, labour and employment law, consumer-protection rules where applicable, data-protection requirements, and protections against unlawful discrimination.
The Code on Social Security, 2020 includes a statutory framework addressing gig workers and platform workers, subject to its commencement and applicable implementation arrangements. The Rajasthan Platform Based Gig Workers (Registration and Welfare) Act, 2023 is also relevant to platform-based gig work in Rajasthan.
However, social-security legislation should not automatically be interpreted as giving every platform worker the same employment status or the same dismissal rights as a conventional employee. The precise rights and remedies depend on the relevant legislation and facts.
A fair platform policy in India should, at a minimum, aim to provide:
Clear information about rating criteria and deactivation grounds.
Accurate records of complaints and performance data.
An opportunity to contest materially incorrect allegations.
A proportionate response to the seriousness of the alleged violation.
A practical grievance mechanism and meaningful review of disputed decisions.
Appropriate safeguards for personal information and confidential records.
These are important compliance and governance practices; whether each is a legally enforceable duty depends on the particular law and circumstances.
7. Practical Example of Deactivation Fairness
Suppose a food-delivery partner receives several low customer ratings in one week. The platform's algorithm automatically deactivates the account, preventing the worker from earning income.
A fair review process should examine:
The reason for low ratings: Were there genuine service failures, or were delays caused by restaurant preparation, traffic, or incorrect customer instructions?
The reliability of evidence: Are the ratings supported by relevant records, or do they contain suspicious patterns?
The applicable policy: Was the worker informed of the rating threshold and the consequences of falling below it?
The opportunity to respond: Can the worker explain the circumstances or challenge inaccurate information?
The proportionality of the decision: Would a warning, temporary restriction, or training be sufficient instead of permanent deactivation?
The appeal mechanism: Can an independent reviewer reconsider the evidence and restore the account if the decision was incorrect?
If credible evidence indicates an immediate safety risk, temporary suspension may be justified while an investigation takes place. However, the final decision should be based on the evidence and applicable legal requirements, rather than solely on an automated score.
8. Best Practices for Digital Platforms
| Area | Recommended practice |
|---|---|
| Rating criteria | Explain the principal factors that affect scores and account eligibility. |
| Data accuracy | Investigate material errors and suspicious rating patterns. |
| Automated decisions | Monitor algorithms for unfair or discriminatory outcomes. |
| Notice | Communicate the reason for deactivation in understandable language. |
| Worker response | Provide a reasonable opportunity to challenge disputed facts. |
| Proportionality | Match the sanction to the seriousness and reliability of the evidence. |
| Appeals | Provide meaningful review and correction of erroneous decisions. |
| Documentation | Maintain records of the decision, evidence, and review outcome. |
| Consistency | Apply platform rules fairly across comparable cases. |
9. Conclusion
Platform ratings and deactivation fairness are central issues in the regulation of the digital economy. Ratings can help maintain service quality, but they can also produce unfair outcomes when they are inaccurate, biased, unexplained, or used without meaningful review.
The seven cases discussed above provide important principles concerning the true nature of platform work, worker protections, contractual arrangements, accessible dispute resolution, and platform regulation. They do not collectively establish a universal legal requirement for a hearing before every deactivation; the applicable obligation depends on the jurisdiction, worker status, contract, and reason for the decision.
The fundamental principle is that a serious deactivation decision should be justified by reliable evidence, applied proportionately, and open to meaningful challenge wherever required by law.

comments