Abstract
A renewal notice seems to ask a narrow question: whether to continue paying. For creative work, that question can carry several others at once. A subscription may provide current access, shared editorial space, support, maintenance, updates, export tools, or a practical coordination habit. A price or renewal date can be clear while the consequence of cancellation, the scope of a download, the continuity of access, the status of shared work, and the burden of migration remain unclear. This paper calls that condition an exit-legibility problem.
Exit legibility is not an accusation against subscriptions, a prediction that a service will fail, a statement of contract law, or a recommendation to leave. It is the degree to which one service relationship makes its payment condition, access gate, data path, cancellation consequence, archive condition, and migration burden understandable enough to appraise. The paper supplies a thirteen-field Creative Service Exit Ledger for an author, editor, small press, program, or tool steward who needs to investigate one actual or contemplated relationship without pretending that a generic policy source can answer a service-specific question. The Ledger distinguishes a fact from an assumption, records the smallest primary evidence needed, and ends with a non-conclusion. It is designed to make a decision more informed, not to make one automatically.
1. A renewal is not a relationship
The renewal notice is short enough to invite a short answer. Cedar Quay Press—a constructed two-person press, not a report about an actual company—opens the notice because one editor is preparing the next season’s copyediting calendar. The message supplies a date, a plan name, and an amount. It also supplies a button that will continue access. On the table beside the laptop are manuscript files, shared comments, a few exported PDFs, a style sheet, a list of freelancers, and a book whose revision has not yet gone to print.
“We own the work,” one editor says. That is true in one ordinary sense: the press created it and has the files it can see. “We pay for the tool,” the other says. That is also true. But neither sentence answers the question the renewal notice has made practical. If the relationship is not renewed, which things continue without interruption, which things stop, which things can be retrieved, which things can be read elsewhere, which people lose a shared way of working, and what would have to be rebuilt?
The press is not trapped. It has not been overcharged, locked out, refused an export, or harmed by a named company. No price, plan, vendor, contract, or cancellation policy is at issue in this scene. The problem is smaller and more common: a visible price has arrived before the relationship surrounding that price has been made legible.
That distinction matters because renewal is a financial event, but a subscription is usually also an access arrangement, a maintenance arrangement, a support arrangement, sometimes a collaboration arrangement, and sometimes a data or file relation. A reader can know the next charge and still not know whether a cancellation changes editing access, what kind of exported material exists, whether another editor’s role changes, whether a history remains readable, or which support function must be replaced. A number is evidence of a number. It is not, on its own, a map of what continues.
A 2024 U.S. Federal Trade Commission announcement about negative-option rulemaking described material terms, informed consent, and a simple cancellation mechanism as distinct elements of recurring subscription practice.[1] The announcement is not a guide to the current enforceability of any rule, a remedy, or a conclusion about a particular creative service. It is useful here because it refuses one common compression: a recurring charge is not the same thing as an intelligible cancellation relation.
Research on cancellation interfaces makes a similar distinction. Sheil and colleagues studied subscription and cancellation flows at popular news websites in four countries. Their study describes barriers in portions of its examined sample, including cancellation flows that required a typed phrase or a call to a representative, and reports inadequate recurring-charge information in many of the subscription flows it examined.[2] The study does not establish that creative software works this way. It does demonstrate why a reader should ask separately about what will be charged, how a relationship can be changed, and what follows after that change.
The International Consumer Protection and Enforcement Network’s 2024 sweep adds a second, broader but still bounded signal. In its stated sample of 642 traders, the report found that, among traders with automatically renewing subscriptions, 81% did not let consumers turn off auto-renewal within the purchase flow; it also reports that 70% lacked information about cancellation steps and 67% lacked a date by which a consumer needed to cancel before another charge.[7] These figures are not a prevalence estimate for creative software and do not identify intent or legality in a named service. They make a teaching point: a subscription relation can contain several independently relevant facts, and the visibility of one is not proof that the others have been answered.
Exit legibility is the degree to which a service relationship makes its ongoing cost, access condition, data path, cancellation consequence, archive condition, and migration burden understandable enough to appraise. It is not a score for a service, a judgment of a contract, or a command to leave.
This paper treats Cedar Quay’s renewal notice as the beginning of an inquiry, not as an answer. Its question is not “Should the press cancel?” It is: What must be shown before the press can tell what it would be renewing, leaving, or replacing?
2. Six verbs, not one ownership word
The word own feels decisive because it seems to settle a relationship. But writers and publishers commonly use it to refer to several different conditions. They may mean authorship of the manuscript, the right to use a service, the possession of a copied file, the ability to download some data, the continued readability of a record, or the ability to end a paid relation. Those are not the same condition.
The following table is not a legal vocabulary and does not interpret a reader’s agreement. It is a language check. Its purpose is to prevent a reassuring word from doing the work of six separate evidence questions.
| Verb | The practical question it opens | What it does not settle |
|---|---|---|
| License | What use is presently authorized under the available evidence? | Whether access persists after payment stops. |
| Access | What account, role, plan, or condition presently permits use? | Whether a readable copy is independently held. |
| Copy | What material exists outside the present service relation? | Whether it preserves every needed relation, comment, version, or workflow. |
| Export | What data or files can be retrieved, in what stated scope and form? | Whether the material will be complete, usable, or accepted elsewhere. |
| Archive | What record is documented as remaining readable and interpretable over time? | Whether a current service account has created that condition. |
| Exit | What changes if the relationship is not renewed, cancelled, or migrated? | Which decision is best for this press. |
Consider how quickly these differences surface at Cedar Quay. The press may own the manuscript in the ordinary creative sense, yet still rely on an account role for shared annotation. It may hold a PDF, yet not know whether a different editing environment can use the comments that influenced the PDF. It may find an export command, yet not know whether the export includes all materials that a specific team member needs. It may know a renewal amount, yet lack a primary, dated description of what cancellation changes.
None of these statements makes the service relation suspect. The distinction is not a trapdoor into a verdict. It is an instruction to name the claim before treating it as settled.
This is particularly important around portability. The European Commission’s Digital Markets Act page describes Article 6(9) in relation to gatekeepers, user authorization, certain data users have provided or generated, and tools or access used for transfer.[3] That scope is useful precisely because it is not universal. It shows that a portability discussion has to ask which data, under which relation, for whom, and through what mechanism.
The Brookings primer on data portability and interoperability makes the conceptual difference explicit. It describes data portability as a user moving data from one company to another and interoperability as systems exchanging data interactively. It also notes that interoperability is commonly partial and that portability can be a one-time download rather than a continuing exchange.[6] The phrase “we can export” therefore does not establish a continuing collaboration relation, an equivalent working environment, or a durable archive. It establishes only what the reader can show about a stated export surface.
The action question is not whether portability is good. It is whether Cedar Quay has named the difference between an export claim and the other claims it is quietly carrying.
3. Hear the relationship, not the interface
A renewal page, an account dashboard, and a support article are interfaces. They can be useful evidence. They are not the entire relationship simply because they are where the reader happens to be looking.
The OECD’s report on dark commercial patterns offers a useful caution here. It defines a policy vocabulary for online choice architecture that can steer, deceive, coerce, or manipulate consumers into choices that may not be in their interests.[4] WEXLEY-WP-043 does not use that vocabulary to label a particular service, interface, or cancellation flow. Ambiguity can arise from a short page, a changed plan, a team’s undocumented practice, a reader’s uncertainty, or a policy that must be read in context. A Ledger is not a dark-pattern detector.
What the OECD source contributes is a modest discipline: visual fluency is not the same as relationship clarity. A smooth button does not prove that a cancellation consequence is known. An easily found export item does not prove that the exported material answers a team’s future needs. A familiar account dashboard does not settle what continues after a plan change. Conversely, a page that requires attention does not prove that a relation is unfair or unworkable.
For this reason, the unit of analysis is not “the software” in the abstract. It is one service relationship. Cedar Quay’s relation might be described this way:
| Relation component | Cedar Quay’s bounded prompt | What should not be assumed |
|---|---|---|
| Work set | Which manuscript, notes, comments, media, or administrative record is actually at issue? | That all press work depends on the same relation. |
| People | Which editors, freelancers, authors, or support roles rely on a shared practice? | That a shared tool is a purely personal purchase. |
| Time | What stated renewal, access, grace, or cancellation timing is actually documented? | That annual payment tells the whole timing story. |
| Service function | What useful work does the service presently support? | That support, maintenance, and coordination have no value. |
| Evidence set | Which dated primary pages, plan records, export descriptions, invoices, or support responses are available? | That memory, a search snippet, or generic policy settles service-specific facts. |
The Ledger does not require the reader to reconstruct a whole technical ecosystem. That would turn this paper into a full publishing-toolchain analysis. Nor does it require a reader to decide whether current account access amounts to custody; that question is treated separately in WEXLEY-WP-024. It asks something more constrained: What relation is this, what material facts are known, and what evidence would make the next decision intelligible?
4. The missing-evidence docket
A service relation becomes harder to appraise when a statement is allowed to slide among three different forms: a verified fact, an assumption, and an unknown. Cedar Quay can use a simple docket to prevent that slide.
| Familiar sentence | What kind of statement might it be? | Evidence question | Unsafe shortcut avoided |
|---|---|---|---|
| “The plan renews annually.” | A dated fact, if tied to the actual plan record. | What source and date state the renewal condition? | Treating a remembered billing rhythm as a full term analysis. |
| “We can cancel whenever we need to.” | An assumption until a primary source defines the relation. | What does a dated source say happens to access, charges, and stored work? | Treating a button label as a complete consequence. |
| “We have the files.” | A partial fact about a specified copy. | Which files, comments, relations, or versions are documented as present? | Treating one visible file as the whole work relation. |
| “We can export everything.” | A scope claim that may need testing or documentation. | What is the documented export surface and what does it omit or leave unknown? | Turning export into archive, interoperability, or substitute workflow. |
| “We could move later.” | A forecast about people, formats, roles, and support. | What work would actually be required, and who would carry it? | Treating movement as frictionless or impossible by definition. |
| “The subscription is too expensive.” | A value judgment that may be important but incomplete. | What price condition is evidenced, and what benefit/dependency is being weighed? | Turning the Ledger into a personal affordability ruling. |
The table does not declare any sentence false. It asks what kind of sentence it is. A reader may discover that a statement is already well evidenced. That is useful. A reader may discover an unknown that requires a primary source. That is also useful. The task is not to make every relation burdensome. It is to prevent an unexamined gap from masquerading as knowledge.
This distinction should be especially strict when a paper might otherwise introduce a product fact. A reader cannot responsibly infer today’s service price, current cancellation condition, plan eligibility, retention period, or export scope from a research source about another service or from a generic policy discussion. The record should instead say:
Evidence rule. Use the exact primary service page, plan record, agreement excerpt, support response, or dated screenshot that establishes the relationship-specific condition. Record the access date and preserve the statement’s scope.
This rule prevents the paper from manufacturing a current product claim merely to make an inquiry appear more complete than its evidence permits.
5. The Creative Service Exit Ledger
The Creative Service Exit Ledger turns the docket into a reader-facing inquiry. It is meant for one bounded relationship: a current subscription, an impending renewal, a contemplated purchase, or a possible migration. It should not be completed as a scorecard. A field may remain unresolved. In many cases, the unresolved field is the Ledger’s most valuable result.
| Field | What to record | Teaching question | Mistake avoided |
|---|---|---|---|
| 1. Work relation | The specific manuscript, title, asset set, or editorial activity actually connected to the service. | What work is this relation supporting? | Treating a whole creative life as one account. |
| 2. Price / renewal condition | A dated, source-named price, renewal cadence, or unresolved price question. | What payment or renewal fact is actually evidenced? | Treating a remembered amount as a complete cost account. |
| 3. Service benefit | A concrete benefit: maintenance, support, collaboration, access, or another stated gain. | What would be lost, changed, or need replacement? | Treating subscription as exploitation or convenience by definition. |
| 4. Access gate | The current credential, plan, role, or condition that permits use. | What condition presently permits access? | Equating a sign-in with ownership or archive custody. |
| 5. Data / export condition | The documented export or retrieval surface, scope, form, date, and unknowns. | What can be retrieved, in what form, under what stated scope? | Treating an export button as proof that all work travels. |
| 6. Cancellation consequence | The primary, dated statement of what cancellation changes, if known. | What does available evidence say stops, continues, or becomes uncertain? | Assuming that “cancel” has a uniform consequence. |
| 7. Archive right or ambiguity | Separate evidence about a retained readable record, archive access, or its absence. | What record, if any, is said to remain legible after service access changes? | Calling any copy, download, or account history an archive automatically. |
| 8. Migration burden | Named work needed to move or substitute the relation: reconfiguration, learning, role changes, compatibility, or unknowns. | What non-price work would change hands? | Reducing a transition to file transfer. |
| 9. Support / maintenance dependency | Support, updates, repair, documentation, or operational assistance tied to the relation. | Which continuing function is being relied on? | Treating access as the only service benefit. |
| 10. Affected collaborators | People, roles, permissions, or shared practices whose relation could change. | Whose work or access may be affected? | Treating a shared relation as private consumption. |
| 11. Policy evidence needed | The smallest missing primary record needed to resolve a material unknown. | What exact page, date, term, or response would clarify this? | Filling uncertainty with a generic rule or assumption. |
| 12. Countervailing benefit | A benefit that makes continuation intelligible even where exit is considered. | What should this record not erase? | Treating scrutiny as proof that exit is best. |
| 13. Non-conclusion | What the Ledger does not decide and the condition that would change its next use. | What remains a choice, not a finding? | Presenting an educational record as a decision engine. |
Cedar Quay might begin only with fields 1, 2, 3, 4, and 11. It can name the manuscript group, cite the renewal notice, explain the current collaborative value, record the access relation, and ask for the smallest missing evidence about cancellation. It does not need to create a speculative migration plan in order to use the Ledger responsibly.
This is why the Ledger stays deliberately bounded. It does not determine where files should live, how backups should be tested, which formats should travel to an archive, or whether an encrypted workspace is appropriate. It gathers the facts that matter when an existing or prospective service relationship must become understandable enough to appraise.
6. What travels, what stops, what changes
A useful exit inquiry separates three sentences that are commonly compressed into one: something travels, something stops, and something changes.
Something may travel. A reader may have documented evidence that a particular file or data set can be retrieved. That can be a meaningful condition. Yet the Brookings distinction remains important: a data transfer may be one-time and partial rather than interactive or continuing.[6] The reader should not assume that a moved object carries every social, technical, or editorial relation associated with the original work.
Something may stop. A current plan could be connected to a role, device, feature, support channel, update stream, account gate, or shared practice. The Ledger does not predict what will stop. It asks for the source that says what the relation itself claims about that condition. If the source is absent or unclear, the result is not “nothing stops.” The result is an unknown that belongs in field 11.
Something may change. A team may need to change its routine even if it can retain files and continue working. Chen and Hitt’s study of online brokerage is relevant in a deliberately limited way. It found that measured switching costs varied in that industry and that service design and usage factors were associated with retention.[5] It does not measure the burden of changing creative software. It is a counterweight to price-only reasoning. The work of transition may include learning, roles, reconfiguration, compatibility checks, support changes, and conversations with collaborators. It may also be manageable. The Ledger does not assume either outcome.
The distinct questions can be held together in one short hearing:
| Hearing question | Evidence category | What the Ledger can record |
|---|---|---|
| What travels? | Primary export/retrieval description, documented file inventory, or stated unknown. | Scope, form, relation, and uncertainty. |
| What stops? | Dated plan, account, cancellation, or support statement. | A stated consequence or an evidence-needed question. |
| What changes? | Bounded description of work, collaborators, maintenance, or support dependencies. | A migration burden to be investigated, not a forecast. |
| What remains valuable? | The reader’s own account of service benefit and countervailing gain. | A benefit that prevents exit from becoming the automatic answer. |
This compact separation produces a more useful action question than “Do we own the tool?” The press may not own the tool in a property sense. That fact alone does not tell it whether to use, renew, or leave the service. What it needs is a readable account of the particular relation it has entered.
7. A proportional evidence request
The Ledger should end in the smallest useful request, not a comprehensive investigation by default. The size of the request should match the consequence of the unknown.
For Cedar Quay, the first question might be simple: What does the current, dated primary material say will happen to the press’s ability to access and retrieve the bounded working materials if this plan is not renewed? That question is narrow enough to be answered by a plan page, account notice, agreement excerpt, support article, or support response. It does not ask a general internet search to answer a service-specific question. It does not ask a policy paper to interpret a contract. And it does not require the press to pretend that an answer is already available.
The evidence hierarchy is intentionally modest:
| Evidence level | Appropriate use | Caution |
|---|---|---|
| Primary, dated service record | Plan terms, account notice, export documentation, cancellation page, agreement, invoice, or written support response. | Record the date, plan relation, and exact scope; a page may still leave relevant questions open. |
| Reader-held operational record | File inventory, team role description, work process note, or documented current practice. | It can describe the reader’s situation but cannot establish a vendor policy. |
| Independent research / policy source | Clarifies concepts such as cancellation mechanism, portability, choice architecture, or switching burden. | It cannot establish the actual terms of the reader’s service relationship. |
| Memory, search result, hearsay, or a generic forum claim | Identifies a question worth checking. | It is not enough to close a material Ledger field. |
This is not a procurement workflow. It does not direct a press to seek legal counsel, negotiate a contract, purchase a replacement, cancel a subscription, or make a particular migration. It is simply a way to prevent a reader from answering a question with the wrong kind of evidence.
The most productive result may be a small sentence: The renewal condition is documented; the cancellation consequence for the team’s relevant work remains unverified; the next needed evidence is a dated primary statement about access and retrieval after cancellation. That sentence is not indecisive. It correctly names the boundary between what is known and what has not yet been shown.
8. Subscription is not the enemy
An exit inquiry can become distorted if it assumes that every recurring service relation is a mistake waiting to be exposed. That is not the premise here.
A subscription may distribute cost over time, fund ongoing maintenance, make timely support available, enable collaboration, reduce up-front barriers, and provide synchronized access that a small press could not otherwise sustain. A team may sensibly decide that the relation’s current benefits exceed the work of a change. It may also decide that an unknown should be clarified before a renewal rather than treated as a reason to leave. The Ledger must make room for both possibilities.
This is why fields 3 and 12 matter. Field 3 records a concrete current service benefit. Field 12 asks what a scrutiny of exit should not erase. A press that lists only dependencies may make its existing work sound irrational. A press that lists only conveniences may make the renewal notice sound self-explanatory. Both records are incomplete.
The value of a service and the legibility of an exit are compatible questions. Better exit legibility does not prove that a reader should exit. It makes continuation more intelligible too. If Cedar Quay can state what it would retain, what it would need to change, and what maintenance/support function it receives, the press can recognize a subscription’s contribution without treating that contribution as a permanent answer.
9. What the Ledger does not decide
The Creative Service Exit Ledger is deliberately narrower than its title can sound. It does not decide whether an account is an archive. WEXLEY-WP-024 distinguishes current account access from warranted claims of custody and archive. WEXLEY-WP-043 records an access gate and an archive ambiguity only because they may matter to a service relationship under appraisal. It does not resolve their preservation meaning.
The Ledger does not trace a return path for a needed earlier state. WEXLEY-WP-027 investigates whether protection, feedback, failure signals, and restore entry make recovery responsibility legible. WEXLEY-WP-043 asks only whether a service relation’s stated cancellation/access condition is known. A reader must not treat an export or a cancellation consequence as proof that recovery is possible.
The Ledger does not calculate the recurring cost of a creative cloud across a working life. That is the economic inquiry reserved to WEXLEY-WP-029. Price and renewal are included here as evidence fields because a reader cannot appraise a service relation without knowing what is being paid or when renewal occurs. The Ledger does not model affordability, future pricing, budget capacity, or life-cycle cost.
It does not assess encryption, keys, permission architectures, or confidentiality trade-offs. Those belong to the collaboration-security inquiry reserved to WEXLEY-WP-030. Nor does it map a title’s distribution, audience, rights, payments, or discovery dependencies; those belong to WEXLEY-WP-041’s publishing-toolchain problem. It does not create a closure protocol for a press ending its activity; that is WEXLEY-WP-044’s organizational continuity question.
Finally, it does not turn the reader’s evidence into a choice. A Ledger cannot know a press’s risk tolerance, mission, finances, collaborators, accessibility needs, creative methods, or future plans. It should not pretend to. The record’s final field is non-conclusion because action can be better informed without becoming automated.
10. What an intelligible exit makes possible
Cedar Quay returns to the renewal notice with more questions than it had at the beginning, but the questions are better shaped. The press can name the relevant work rather than saying “our whole archive.” It can distinguish a price line from a cancellation consequence. It can recognize shared editorial practices as part of a relation without calling those practices a form of ownership. It can ask what a particular export is documented to contain without claiming that every export is a substitute for an archive. It can retain the benefits of maintenance and support in the same record that names an unresolved access condition.
This is the point of exit legibility. It does not make a reader suspicious by default. It makes a reader specific. Instead of asking whether a writer “owns” a tool, the Ledger asks what the reader can actually show about a bounded service relation: the stated price, current access gate, data/export surface, cancellation consequence, archive ambiguity, migration work, support dependency, collaborator effect, and missing evidence.
The next act might be to continue, to ask a support question, to locate an export description, to read a plan record, to prepare a future alternative, or to decide later. This paper does not choose among those acts. Its final question is smaller and more demanding:
What continues, what stops, what travels, and what must be shown before this service relationship is renewed, left, or changed?
References
[1] U.S. Federal Trade Commission. “Federal Trade Commission Announces Final ‘Click-to-Cancel’ Rule Making It Easier for Consumers to End Recurring Subscriptions and Memberships.” 16 October 2024. https://www.ftc.gov/news-events/news/press-releases/2024/10/federal-trade-commission-announces-final-click-cancel-rule-making-it-easier-consumers-end-recurring.
[2] Sheil, A., Acar, G., Schraffenberger, H., Gellert, R., and colleagues. “Staying at the Roach Motel: Cross-Country Analysis of Manipulative Subscription and Cancellation Flows.” Proceedings of the CHI Conference on Human Factors in Computing Systems, 2024. https://doi.org/10.1145/3613904.3642881. Open manuscript: https://arxiv.org/abs/2309.17145.
[3] European Commission, Digital Markets Act. “End User Data Portability.” Accessed 21 August 2026. https://digital-markets-act.ec.europa.eu/developer-portal/end-user-data-portability_en.
[4] Organisation for Economic Co-operation and Development. Dark Commercial Patterns. OECD Digital Economy Papers, 2022. https://www.oecd.org/en/publications/dark-commercial-patterns_44f5e846-en.html.
[5] Chen, P. Y., and Hitt, L. M. “Measuring Switching Costs and the Determinants of Customer Retention in Internet-Enabled Businesses: A Study of the Online Brokerage Industry.” Information Systems Research 13, no. 3 (2002): 255–274. https://doi.org/10.1287/isre.13.3.255.78. Prepublication manuscript: https://oz.stern.nyu.edu/seminar/sp02/0204.pdf.
[6] Brookings Institution. “Data Portability and Interoperability: A Primer on Two Policy Tools for Regulation of Digitized Industries.” 9 May 2023. https://www.brookings.edu/articles/data-portability-and-interoperability-a-primer-on-two-policy-tools-for-regulation-of-digitized-industries-2/.
[7] International Consumer Protection and Enforcement Network. ICPEN Dark Patterns in Subscription Services Sweep: Public Report 2024. 2024. https://www.oaic.gov.au/__data/assets/pdf_file/0018/232623/Public-Report-ICPEN-Dark-Patterns-Sweep.pdf.