sended
Back to Home

Acceptable Use Policy

Version: 2.1-draft · Effective date: [not yet in effect]

This document is a draft under legal review. It is not yet published and not yet in effect.

1. Introduction and Purpose

1.1. This Acceptable Use Policy ("Policy") is issued by Elephants Production s.r.o., with its registered seat at Rybná 716/24, Staré Město, 110 00 Prague 1, Czech Republic, Company ID (IČO): 23397292, registered in the Commercial Register maintained by the Municipal Court in Prague, File No. C 426391 (the "Provider"), and governs the use of the Sended platform at sended.io ("Sended" or the "Service").

1.2. This Policy forms an integral part of, and is incorporated by reference into, the Terms of Service (terms-of-service.md). It is contract document 4 in the order of precedence set out in Section 3.4 of the Terms of Service; it is presented to the Customer for acceptance at registration alongside the Terms of Service and the Privacy Policy, and the Provider records which version of it the Customer accepted and when. Capitalized terms not defined here have the meaning given in the Terms of Service. In particular: "Customer" means the salon business holding the account (the Service is offered to businesses only), "Authorized User" means an individual using the account, and "End Client" means the Customer's client or prospective client whose personal data is processed through the Service on the Customer's behalf.

1.3. Why platform-level rules exist. Sended delivers messages through shared sending infrastructure: a single Amazon SES account and shared sending domains (whose IP and domain reputation is common to all Customers), and the Provider's integration with Meta's WhatsApp Business Platform. Mailbox providers and messaging platforms measure the behaviour of this shared infrastructure as a whole. One Customer's spam, high bounce rates, or policy violations therefore degrade deliverability — or trigger enforcement by Amazon or Meta — for every Customer on the platform. This Policy exists to protect the shared infrastructure, all Customers, and the recipients of messages. The Provider enforces it accordingly, including through the automated safeguards and the enforcement measures described below.

1.4. This Policy supplements, and does not replace, the Customer's own legal obligations under the GDPR, Czech Act No. 480/2004 Coll., on certain information society services, and any other law applicable to the Customer's communications. Compliance with this Policy does not by itself constitute compliance with the law.

1.5. Consumers. The Service is offered to businesses only, and the Customer confirms at registration that it acts within the scope of its business activity and not as a consumer (Terms of Service, Section 3.1). Where mandatory law nevertheless treats a particular Customer as a consumer, this Policy applies to that Customer only to the extent it does not conflict with mandatory provisions protecting consumers.

2. General Principle: a Lawful Basis for Every Message

2.1. The Customer may use the Service to send marketing messages only to End Clients for whom the Customer has a lawful basis, meaning either:

  • (a) the End Client's prior consent to receiving marketing communications from the Customer; or
  • (b) the Czech "existing-customer" exception (soft opt-in) under Section 7(3) of Act No. 480/2004 Coll., which applies only where all of the following are true:
    • the Customer obtained the electronic contact details from that End Client itself, in connection with the sale of the Customer's own product or service;
    • the messages promote only the Customer's own similar products or services;
    • the End Client had the opportunity to refuse such use of their contact details when they were collected; and
    • every message offers a simple, free-of-charge way to refuse further messages.

2.2. Promotion of third-party products or services always requires the End Client's consent; the soft opt-in does not cover it.

2.3. The Customer is solely responsible for being able to demonstrate the lawful basis for each recipient. The allocation of data protection roles and the parties' related obligations are set out in the Data Processing Agreement (data-processing-agreement.md) and the Privacy Policy (privacy-policy.md).

3. Email Rules

3.1. Safeguards the Service applies automatically

The Service enforces the following on every marketing email. These safeguards are designed to support the Customer in meeting the labeling and opt-out duties of Act No. 480/2004 Coll. and the one-click unsubscribe requirements of major mailbox providers; they do not by themselves discharge those duties. The system footer is rendered in the language the Customer selects for its workspace (English, Czech, Russian, Slovak or Ukrainian); where the Customer's recipients require another language, the Customer remains responsible for meeting the labeling and refusal requirements that apply to its communications.

  • (a) Commercial-communication labeling. Every marketing email carries a clear commercial-communication marker in a system-owned footer that the Customer cannot edit or remove.
  • (b) Sender identity. The footer identifies the Customer by name, Company ID (IČO), and registered seat. If the Customer has not provided its IČO and registered seat, the email channel remains closed and marketing email cannot be sent.
  • (c) One-click unsubscribe, enforced across every channel. Every marketing email contains a per-recipient unsubscribe link and the corresponding one-click unsubscribe headers (RFC 8058). An opt-out recorded through that link, from a WhatsApp stop signal (Meta's user-preference signal for that recipient, or a delivery error reporting the same), or by the salon itself for a refusal it received elsewhere, takes effect for all marketing channels — email, WhatsApp, and Telegram — not only the channel on which it was made. There is no automatic stop detection on Telegram, and a refusal expressed as a free-text reply on WhatsApp is not detected either: those the Customer must record in the Service, which then enforces them on every channel (Sections 5.7 and 6.4). Each opt-out is scoped to the Customer whose message it was: the objection is to that salon's direct marketing (GDPR Art 21(3); Act 480/2004 §7 concerns the sender of the communication). A person who is a client of several salons that use the Service opts out of each salon separately, and the unsubscribe page names the salon concerned so the scope is clear to the recipient. One salon's opt-out never blocks another salon's lawful marketing to the same person. A recorded objection can be withdrawn only by the recipient, and only by an explicit action on the email unsubscribe page reached from that salon's unsubscribe link; an opt-back-in on WhatsApp does not remove it, and the Customer cannot remove it at all.
  • (d) Global suppression list. In contrast to the per-salon opt-out in (c), email addresses that hard-bounce or raise a spam complaint are suppressed platform-wide, across all Customers. A complaint or a dead address is a signal about the Provider's shared sending infrastructure, not an objection to one salon: mailbox providers rate the sending domain as a whole, and continuing to mail such an address on behalf of any Customer would endanger deliverability for every Customer. Suppressed addresses are not mailed again; a suppression can be lifted only by the Provider, for a documented reason.
  • (e) Frequency cap. At most one marketing message per client record per 24 hours is delivered on behalf of a Customer, across campaigns and marketing automations. The cap's identity is the client record in the Customer's booking system, not the person behind it: where that system holds more than one record for the same individual, each record carries its own cap and that individual may receive one message per record.

Transactional messages (e.g. appointment reminders) are exempt from marketing-specific safeguards such as the frequency cap, but remain subject to the global suppression list. The Customer should note that automation emails currently carry the same system footer whatever their purpose: an appointment reminder therefore also displays the commercial-communication marker and an unsubscribe link. An unsubscribe clicked on such a message records a marketing opt-out only — appointment and other transactional messages continue to be sent — so the recipient's action is narrower than the footer suggests to them. Until the footer distinguishes the two cases, the Customer must take this into account when it configures transactional automations and when it handles queries from recipients about it.

3.1a. Configurable controls the Service offers

Quiet hours. In addition to the safeguards in Section 3.1, the Service offers quiet hours: where the Customer enables them, scheduled marketing campaign and marketing automation sends are deferred until the window ends (evaluated in the timezone configured for the Customer's account), not dropped. Quiet hours are off by default — a Customer that never opens the setting has none — they apply to marketing sends only, transactional messages such as appointment reminders being exempt, and the Customer may override them for an individual campaign. Quiet hours are a convenience control, not a compliance guarantee: the Customer remains responsible for observing any legal or platform restriction on the times at which its communications may be sent (Terms of Service, Section 10.1a).

3.2. Customer obligations and prohibitions

The Customer must not circumvent, and must not attempt to circumvent, the safeguards in Section 3.1. Using a control the Service offers the Customer — in particular leaving quiet hours off, or overriding them for an individual campaign under Section 3.1a — is not circumvention. In particular, the Customer must not:

  • (a) use purchased, rented, harvested, or otherwise third-party-sourced contact lists. This is prohibited outright: the soft opt-in under Section 2.1(b) can never apply to contacts the Customer did not collect from its own customers, and importing such lists into the Service is a material breach of this Policy regardless of any consent the list vendor claims to hold;
  • (b) use misleading or deceptive subject lines, disguise the commercial nature of a message, or conceal or misrepresent the sender's identity;
  • (c) re-add, re-import, or otherwise attempt to message End Clients who have opted out, or alter contact details (e.g. modified email addresses) to evade the suppression list or an opt-out;
  • (d) interfere with the unsubscribe link, the unsubscribe headers, or the system footer, or instruct End Clients to ignore them;
  • (e) provide false or inaccurate company identity details (IČO, registered seat) for the sender footer.

The Customer must keep its company identity details accurate and up to date, and must honour opt-out requests it receives outside the Service (e.g. by reply email or in person) by not messaging those End Clients through the Service.

The duty also runs the other way. An opt-out recorded in the Service is the End Client's objection to the Customer's direct marketing as such, not merely a setting in this platform: the Customer must give effect to it in every system it uses to contact that End Client, and must not export, copy, or otherwise take contact data out of the Service in order to send that End Client messages the Service would have refused to send.

4. Deliverability and Shared Reputation

4.1. Amazon SES, through which all email is sent, enforces account-level reputation thresholds — indicatively, a bounce rate around 5% places the sending account under review, and a complaint rate should remain below approximately 0.1%. Because these thresholds apply to the Provider's shared account, a single Customer's poor list quality endangers email delivery for all Customers.

4.2. The Service continuously monitors each Customer's email deliverability: it computes rolling bounce and complaint rates per Customer from delivery events (once a Customer's volume is large enough for the rates to be meaningful) and maintains a service-wide suppression list — an address that hard-bounces or complains is automatically excluded from further marketing email across the Service.

Because Amazon reports feedback by recipient address and not by sending workspace, a bounce or complaint for an address is attributed to every Customer that successfully emailed that address within the same measurement window. An address that two salons both mailed that week contributes its bounce to both of them. Sharing recipients with another Customer whose list quality is poor can therefore affect the Customer's own measured rates.

The Service warns the Customer in the application at rates set below Amazon's account-level thresholds — currently a bounce rate of 3% or a complaint rate of 0.05% — and automatically pauses that Customer's marketing email at Amazon's own thresholds, currently a bounce rate of 5% or a complaint rate of 0.1% (other channels and transactional messages are unaffected). The Customer is notified of each change.

The pause lifts automatically once the Customer's rates have stayed below the pause levels for a full measurement window. Lifting the pause does not necessarily restore the normal state: a Customer whose rates are still at or above the warning levels returns to the warned state, and can remain there while it continues to send. Where the Customer's volume is too low for the rates to be measured at all, a full measurement window with no bounces and no complaints likewise lifts the pause. The Provider can also lift a pause manually, for a documented reason. Where the risk to the platform is acute, the Provider may additionally pause or throttle sending directly, with notice given promptly.

4.3. The Customer must maintain reasonable list hygiene: sending only to contacts collected as described in Section 2, and not repeatedly attempting delivery to addresses that bounce.

5. WhatsApp Rules

5.1. WhatsApp messages are sent via Meta's WhatsApp Business Platform (Cloud API). The Customer's use of the WhatsApp channel is additionally subject to Meta's WhatsApp Business Terms of Service and the WhatsApp Business Messaging Policy, which the Customer must comply with.

5.2. Templates. Business-initiated messages (campaigns and automations) may be sent only using message templates approved by Meta, in the correct template category. The Service enforces this: business-initiated WhatsApp sends without an approved template are not dispatched.

5.3. Opt-in. Under the WhatsApp Business Messaging Policy, the Customer may send WhatsApp messages only to persons who have provided their phone number to the Customer and given opt-in permission to be contacted by the Customer on WhatsApp. The Customer is solely responsible for obtaining opt-in by a lawful method and must honour opt-out requests whether made on or off WhatsApp. The Service supports this: it keeps a record of WhatsApp permission events (opt-ins, opt-outs, and their evidence) per End Client, and dispatches WhatsApp messages only to End Clients for whom a recorded opt-in exists. The Customer must therefore record its opt-ins, and any withdrawals, in the Service: an End Client with a recorded opt-out is not messaged, and neither is one for whom no permission record exists at all.

5.4. 24-hour customer-service window. Free-form (non-template) messages may be sent only in reply to an End Client's message, within the 24-hour customer-service window defined by Meta. The Service checks the window before dispatching free-form replies; outside the window, only approved templates may be used. Separately, Meta requires an active payment method on the Customer's WhatsApp Business Account for certain conversation categories (per Meta's pricing policy, from 1 October 2026); where the Customer's Meta payment setup is not ready, the Service refuses to dispatch the affected replies until it is (see Terms of Service, Section 7.5).

5.5. Prohibited content. The Customer must not use the WhatsApp channel for content prohibited by the WhatsApp Business Messaging Policy, including its restricted business verticals (e.g. drugs, weapons, adult content) and country-specific restrictions.

5.6. Meta enforcement. Meta applies its own quality-rating enforcement (messaging limits, template pausing, temporary or permanent blocks) at the level of the business account. The Customer acknowledges that Meta's enforcement actions are outside the Provider's control and that repeated poor quality ratings may result in the Provider suspending the Customer's WhatsApp channel under Section 10.

5.7. Commercial-communication labeling and opt-out. Unlike marketing email (Section 3.1), WhatsApp messages do not carry a system-inserted commercial-communication marker or opt-out mechanism. The Customer must therefore include in every WhatsApp marketing template: (i) a clear identification of the message as a commercial communication, and (ii) a working, free-of-charge opt-out instruction (e.g. an invitation to reply "STOP"). The Customer must honour such replies and other opt-out requests without undue delay, so that no further marketing is sent to that End Client. Where Meta reports a stop for a recipient — through its user-preference signal or a delivery error to the same effect — the Service records the objection itself and enforces it across all marketing channels (Section 3.1(c)). A refusal that reaches the Customer only as a free-text reply is not detected automatically; the Customer must record it in the Service. Where the conditions of the soft opt-in under Section 2.1(b) are not clearly met, the Customer must treat WhatsApp marketing as consent-only.

6. Telegram Rules

6.1. The Telegram channel operates through the Customer's own connected Telegram account — messages are sent from the Customer's account, under the Customer's identity, not from a bot or an account of the Provider.

6.2. Telegram's Terms of Service prohibit spam and unsolicited messaging. Because sends originate from the Customer's own account, the Customer bears the consequences of violating Telegram's terms — including limitation, restriction, or permanent ban of the Customer's Telegram account by Telegram. The Provider has no ability to prevent or reverse such measures.

6.3. The Customer must therefore message via Telegram only End Clients covered by a lawful basis under Section 2, and should exercise particular restraint on this channel. The Provider's platform safeguards — the channel-wide marketing opt-out and the frequency cap — apply to Telegram sends in the same way as to other channels, and the Customer must not attempt to circumvent them. Quiet hours, where the Customer has enabled them (Section 3.1a), apply to scheduled marketing sends on this channel too.

6.4. Commercial-communication labeling and opt-out. Unlike marketing email (Section 3.1), Telegram messages do not carry a system-inserted commercial-communication marker or opt-out mechanism. The Customer must therefore include in every Telegram marketing message: (i) a clear identification of the message as a commercial communication, and (ii) a working, free-of-charge opt-out instruction (e.g. an invitation to reply "STOP"). The Service does not read inbound Telegram messages for stop instructions: a refusal received on this channel is not detected automatically. The Customer must honour such replies and other opt-out requests without undue delay and must record them in the Service, which then enforces the objection across all marketing channels (Section 3.1(c)), so that no further marketing is sent to that End Client. Where the conditions of the soft opt-in under Section 2.1(b) are not clearly met, the Customer must treat Telegram marketing as consent-only.

7. Instagram Direct

Instagram Direct messaging is in development and available only where and when enabled for the Customer's account. Where enabled, its use is subject to this Policy and to Meta's applicable platform terms and policies, on the same principles as Section 5.

8. Prohibited Content (All Channels)

The Customer must not use the Service to create or send content that:

  • (a) is illegal under the law applicable to the Customer or the recipient, or promotes illegal activity;
  • (b) is fraudulent, deceptive, or misleading, including impersonating another person or business;
  • (c) is defamatory, harassing, threatening, or incites hatred or violence;
  • (d) contains malware, phishing links, or other technically harmful material;
  • (e) sexually exploits or endangers minors in any way;
  • (f) violates the intellectual property or personality rights of others; or
  • (g) violates the content policies of the messaging platform used for delivery (see Sections 5–7).

Additionally, images uploaded to the Service for use in messages and campaigns are distributed through a public content-delivery network under a long-lived public URL (see the Sub-processor List, Section 3.1). The Customer must not upload images containing End Client personal data — in particular photographs of its clients — or any other content it is not entitled to publish.

An uploaded image cannot be withdrawn. Removing an image from the Service's library hides it from further use but does not delete the stored object or its public URL, which remains reachable and may stay cached at content-delivery edges for up to one year; messages already sent continue to reference it, and the object is retained after termination of the Agreement for that reason. The Customer must therefore treat every upload as a publication (Terms of Service, Section 14.5).

9. AI Feature Rules

Where AI-assisted features are enabled for the Customer's account (e.g. AI-drafted replies to End Client messages and AI-assisted automations), the following applies:

  • (a) Disclosures stay intact. The Customer must not remove, suppress, or instruct Authorized Users or End Clients to disregard any AI-interaction disclosure that the Service displays or inserts, and must not present AI-generated interactions to End Clients as exclusively human where applicable law (including Article 50 of the EU AI Act, Regulation (EU) 2024/1689) requires transparency.
  • (b) Review where configured. In the default, semi-automatic mode, the AI assistant sends conversational replies to End Clients autonomously; what the Service presents for approval are proposed actions in the Customer's booking system (creating or cancelling appointments, creating client records). The Customer must actually review such proposals before approving them, and must monitor AI-assisted conversations with reasonable regularity. Approval without review does not shift responsibility to the Provider.
  • (c) No misleading output. The Customer must not use AI features to generate content that is misleading, deceptive, or otherwise prohibited under Section 8, and must not attempt to prompt the AI features into producing such content.
  • (d) Responsibility. All messages sent from the Customer's account — whether drafted by a human or with AI assistance — are the Customer's communications, and the Customer remains fully responsible for them.

10. Enforcement

10.1. Where the Provider determines, acting reasonably, that a Customer has breached this Policy or that the Customer's sending endangers the platform, the Provider may apply the following measures, escalating as appropriate to the severity, duration, and recurrence of the issue:

  1. Warning — notice of the breach and the required correction;
  2. Throttling — reduction of the Customer's sending rate or volume;
  3. Channel suspension — suspension of the affected channel (e.g. email or WhatsApp) while other channels remain available;
  4. Account suspension — suspension of the Customer's sending capabilities or of the account, in accordance with the Terms of Service.

10.2. The Provider is not required to follow the ladder sequentially. For severe abuse — including use of purchased lists, deliberate circumvention of opt-outs or suppression, illegal content, phishing, or conduct that triggers or imminently risks enforcement action by Amazon, Meta, or Telegram against the Provider's infrastructure — the Provider may suspend the affected channel or the account immediately, with notice given promptly afterwards. Section 18.1 of the Terms of Service carves the conduct described in this Section out of the notice-and-cure requirement that otherwise applies before the Provider suspends for a curable breach.

10.3. Suspension under this Policy does not affect Fees already due or obligations already accrued. It does not affect the Customer's right to terminate the Agreement at any time (Terms of Service, Section 19.2), and the Provider lifts a measure once its cause is remedied (Terms of Service, Section 10.3).

10.4. Appeal. The Customer may contest an enforcement measure by writing to contact@sended.io, describing the measure and the grounds for review. The Provider will review the appeal and respond without undue delay; the Provider will lift or reduce the measure where the review shows it is no longer warranted.

10.5. Nothing in this Section affects the rights of a Customer that mandatory law treats as a consumer (Terms of Service, Section 3.1).

11. Reporting Abuse

Anyone may report suspected abuse of the Service (spam, unlawful content, impersonation) to contact@sended.io. Requests concerning personal data (including opt-out and erasure requests by End Clients) may also be directed to privacy@sended.io and are handled as described in the Privacy Policy (privacy-policy.md).

12. Changes to this Policy

The Provider may update this Policy as the Service, the applicable law, or the requirements of the underlying messaging platforms evolve. Section 21 of the Terms of Service governs amendments to this Policy as it governs amendments to the Terms themselves: the Provider notifies the Customer at least 30 days before an amendment takes effect, and the Customer may reject the amendment and terminate as set out there. The version and effective date of this Policy are stated at the top of this document.

13. Governing Law

This Policy is governed by the laws of the Czech Republic. Any disputes arising from or in connection with this Policy fall within the jurisdiction of the courts of Prague, Czech Republic, as further set out in the Terms of Service. This does not affect the rights of a Customer that mandatory law treats as a consumer (Terms of Service, Section 3.1), including the right to use the out-of-court dispute resolution route identified in Section 22.3 of the Terms of Service.