Privacy policy

Who is responsible

SolidQR, LLC, an Arkansas limited liability company doing business as Graklink, operates graklink.com and the services on grakl.ink, grak.link and graklink.net. Contact support@graklink.com or use our contact form for privacy questions or requests.

We decide how account, billing, security and service-administration information is used. Customers choose their destinations, hosted content and measurement configuration and may have their own responsibilities for visitor data. A customer's destination website has its own privacy practices; this policy does not replace its notice. If your request concerns that business's use of information, contact it or give us the affected Graklink address so we can help route the request.

Information and purposes

We receive information directly from you when you create an account, configure a product or contact us; from browsers and devices when someone uses our services or follows a Graklink address; and from providers that process payments, verify requests or supply a connection you authorize. Customers can also submit information through their content, settings and reports.

You can read our public pages without creating an account. An email address is needed for account access and replies to support requests, and payment information is needed for a purchase. Recovery email, business context and Google Analytics connection are optional. Leaving an optional feature unconfigured limits that feature, rather than preventing ordinary use of your QR codes and short links.

• Account and purchases: email, optional recovery email, authentication records, product ownership, orders, subscription state, payment references, communication preferences, and guided-introduction settings and progress. We use these to provide access, fulfill purchases, prevent abuse and communicate about the service. Stripe processes payment-card details; Graklink does not receive your full card number or security code.
• QR scans, short-link clicks and hosted-page activity: timestamps, item and routing identifiers, approximate country/region/city, device/browser/operating-system categories, browser user-agent strings (up to 300 characters), browser language preferences (up to 60 characters from the Accept-Language header), referring source and available request-quality signals. An IP address can be used to derive location and a keyed network identifier for audience estimates; raw IP addresses are not stored in these analytics records. We also record visits to hosted pages and cards and, when activity measurement is enabled, actions such as contact-link clicks and downloads. A marked contact click records that action, not proof of a completed enquiry or sale. Network identifiers are pseudonymous, not proof of a person's or device's identity. Security and infrastructure processing can handle IP addresses separately.
• Product configuration and uploads: destinations, routing and tracking settings, item names, campaign details, design files, logos, published pages and cards. We store and process these to operate what you create. Uploaded images may be sent to Sightengine for screening, including queued screening when that service is unavailable. The image itself can contain personal information; do not upload material you lack permission to share.
• Goals and reporting: business context you confirm, goals, campaign costs you enter, report requests, published reports and feedback on recommendations. These help provide analysis and follow-up. Do not include information about identifiable visitors or sensitive personal details in these fields.
• Contact, Support and feedback: reply email, submitted text, ticket history, attachments, verification records and delivery metadata. We use them to respond, resolve issues and consider product suggestions. Staff notes remain internal. Abuse reports may require sharing relevant allegations with the affected party; do not submit unnecessary private information.
• Availability requests: your email or account, requested address, and any affiliate-program region you provide. We use these for the requested updates. Each availability email offers withdrawal from that request; it is not newsletter consent.
• Website use and security: aggregate counts for actions such as pricing visits, article engagement and builder use, plus request information needed for rate limits, fraud prevention and debugging. Marketing-event storage uses approved event/page labels rather than arbitrary submitted text. Infrastructure logs and anti-abuse services can process IP addresses, headers and browser signals.
• Affiliate applications, when the program is available: name, email, application and decision records, payout handle, referrals and commission records. Required tax information is encrypted and retained when needed for tax obligations. Program availability and eligibility are separate from joining its waitlist.

Cookies and browser storage

Browser storage supports sign-in, purchases, saved interface state, security and the optional measurement described below. Blocking it can prevent sign-in, cart recovery or a configured measurement feature from working.

• Sign-in: a dashboard session cookie normally lasts 30 days unless revoked earlier. Short-lived pending-verification state supports sign-in. Guest Support uses a separate verified-session cookie. Signing out removes the corresponding session; Sign out everywhere revokes dashboard sessions.
• Cart and reservations: cart and cart-count cookies can last 30 days. Temporary address holds expire after a few minutes. Browser storage can also retain form drafts, interface selections or demo introduction progress so you can continue a task. Demo introduction progress is separate from account preferences.
• Website measurement: session storage remembers whether an engagement or pricing event was already counted in the current browser session. These flags are distinct from authentication cookies and do not contain a visitor profile.
• Security verification: Cloudflare Turnstile checks selected public requests, such as sign-in, contact, waitlist and anonymous checkout submissions. Cloudflare processes browser and request signals to distinguish legitimate requests from automated abuse. Ordinary browsing and QR/short-link redirects do not use this form-verification flow. See Cloudflare's privacy policy.
• Affiliate attribution, only when enabled: a referral link can set a first-party referral cookie for up to 90 days. It records the affiliate code and may be associated with a purchase and commission. The affiliate program and its attribution are not activated simply by joining a waitlist.

We do not treat a cookie as exempt from consent merely because it is first-party. The cookie settings banner stores your website category choices in a first-party preference cookie without a visitor identifier. The optional experiment controls described below apply to a separate measurement path. Browser controls can clear storage; doing so also removes remembered sign-in and measurement choices. Do Not Track is not a service-wide setting. It blocks our website analytics until you explicitly enable Analytics in cookie settings. That choice allows Google Analytics, while Umami stays off. The optional experiment path honors Global Privacy Control by excluding identifier-based measurement.

Website analytics

On our own public website, the Analytics cookie setting controls Google Analytics 4. We load its tag only after you enable that setting. If you leave it off or have not chosen yet, the existing Umami website measurement remains in use unless your browser privacy signal or the separate website analytics opt-out below blocks it. Only one of these two website analytics tools sends events at a time. We host the Umami database ourselves; Cloudflare delivers its tracking requests and protects access to the reports. The Marketing setting controls Google's advertising consent signals separately from Analytics. The Personalization setting stores that choice; it does not turn on a personalization feature by itself.

Our website event routing sends approved page paths, referring website domains, selected campaign tags and interaction names. We exclude form contents, customer destinations, sign-in tokens and unrelated URL parameters from events we compose. Google Analytics may use its own browser storage and device information after you enable Analytics. Umami processes the request's IP address and browser information to derive a rotating visitor identifier; its analytics records do not store the raw IP address. We do not collect Umami session replays or heatmaps. Global Privacy Control, Do Not Track and the separate website analytics opt-out below stop Umami website tracking.

We also use Grakel, a cookieless visit counter operated by our affiliated product Grakel, on public website pages. It receives the page address without its query string or fragment (except campaign tags), the referring website's domain when you arrive from another site, and browser information per pageview. It sets no cookies or browser storage and does not store IP addresses. Global Privacy Control, Do Not Track and the website analytics opt-out below stop it. Enabling Analytics for Google Analytics does not turn Grakel off.

When you proceed from the public cart to checkout, we can record a hashed checkout reference and match it to a confirmed order. Our conversion reports use paid-order amounts and keep purchases without a matching tracked visit unattributed. They do not add your name, email address or payment details to website analytics.

We retain Umami website analytics events for up to 13 months. Encrypted backups expire separately under our backup retention schedule. Umami does not set a visitor-identification cookie. Google Analytics has its own storage and retention settings. Changing the Analytics cookie setting stops new events in the previous website analytics tool; it does not delete earlier events already held there. These controls apply to this website analytics feature; service operation, security records and the other measurement paths described in this policy have their own rules.

We also use Cloudflare Web Analytics to measure page views and browser performance, including page load times. Its performance beacon runs separately from the Graklink website analytics choice below. See Cloudflare Web Analytics for details about this service.

Optional experiment measurement

Ordinary scan/click counting does not require a visitor tracking cookie or an interruption to the redirect. A customer can separately configure an eligible outcome experiment. Where its measurement policy permits, a signed first-party cookie stores a random identifier and measurement choice scoped to that customer and measurement period. It is not a universal identity shared across Graklink customers. The identifier expires at the end of the configured daily, monthly or quarterly period, no later than 92 days.

Where valid consent is required and unavailable, identifier-based experiment measurement is excluded. A customer may explicitly opt into a redirect consent dialog for supported jurisdictions. The visitor can allow measurement or continue without it; either choice reaches the destination. The dialog is not the default redirect behavior. Unknown or unsupported policy states do not establish permission to measure.

For an eligible measured visit, Graklink can send a short-lived signed receipt to the configured destination and accept a validated outcome linked to that receipt. Assignment, variant, configuration version, outcome and deduplication records support the comparison. This connects an eligible browser visit with an outcome for that experiment; it does not identify a known person or prove that a QR code caused a sale.

Use the experiment's measurement-preferences link to stop identifier-based measurement in that browser. The supported withdrawal flow removes that browser's current experiment outcomes from the comparison. You can also contact us with the affected address. We cannot reliably identify a past anonymous visitor from a name or email we never collected for that visit.

Google Analytics and AI

Google Analytics 4 is an optional integration. New connections are currently paused pending Google review. The following describes how the integration handles data when enabled. After your authorization, we access available account/property information and read reporting data for the property you select using the analytics.readonly permission. Reporting data can include dates, campaign names and identifiers, source and medium, location and device categories, sessions, users, key events, purchases and revenue. We store encrypted authorization credentials, selected property and timezone/currency settings, sync status and imported reporting aggregates. We use them to show website results alongside your Graklink activity and support reports and recommendations. We do not modify your Google Analytics property or use Google data for advertising.

You can disconnect in the dashboard or revoke access in your Google account. Revocation stops future authorized reads; use Graklink's disconnect/delete action or contact us to remove imported connection data. That action starts removal of credentials and imported data, with the dashboard showing its status. If Google revocation cannot be confirmed, you can also revoke access directly in your Google account. Published reports or exported files can contain historical values; copies already downloaded or sent outside Graklink cannot be recalled.

AI-assisted insights use Anthropic's Claude models, reached through Cloudflare AI Gateway, to process selected aggregate evidence and confirmed business context. We check analytical inputs for common direct identifiers, exclude certain low-volume counts and validate returned selections before presenting recommendations. These checks reduce exposure; they do not guarantee that every submitted value is anonymous. Selected Google reporting aggregates can be included when they support those features. We configure these AI requests to bypass gateway caching and payload logging; reports and operational metadata are stored separately. AI can still be wrong; the dashboard provides the underlying evidence. We do not use customer analytics or Google data to train general-purpose models. Anthropic states that by default it does not use inputs or outputs from its commercial products to train its models; see Anthropic's model-training statement.

Our use and transfer of information received from Google APIs is limited to the disclosed user-facing features and permitted security or legal purposes, in accordance with the Google API Services User Data Policy, including applicable Limited Use requirements.

How we protect information

We use authentication and account-scoped access checks to restrict access to private account information. Stored Google Analytics refresh credentials are encrypted using AES-256-GCM. The encryption binds each credential to its account and connection, and the encryption key is supplied separately from the stored credential record. Requests to Google's authorization and Analytics services use HTTPS.

We limit the information included in analytical processing. Raw IP addresses are not stored in QR scan, short-link click or hosted-page analytics records, and AI input checks are described above. This does not mean that all service records are anonymous: support messages, account records and some operational logs can contain personal information. No storage or transmission method can guarantee absolute security. If you suspect unauthorized access or a privacy incident, contact support@graklink.com.

Uses we exclude

We do not sell or rent personal information, create advertising audiences from customer analytics, or share personal information for cross-context behavioral advertising. Customer analytics are not published for other customers. We do not provide a list of identified scanners. These commitments do not mean information is never processed by service providers, included in your exports or disclosed for a valid legal purpose.

Providers and disclosure

Providers receive information needed for their role: Cloudflare for hosting, storage, delivery, security and routing AI requests; Anthropic for AI-assisted insights; Stripe for payments and subscription administration; Resend and configured mail services for customer and Support email; Sightengine for uploaded-image screening; and Google for authorized Analytics access and destination-threat checks. Destination checking may send the URL being checked to a safety service. Do not put secrets or personal details in URLs.

We may disclose information to authorized staff and providers to deliver support, investigate fraud or abuse, protect rights, or comply with applicable law and valid legal process. We give notice of legal demands when permitted and appropriate. You control exports, printed reports, downloaded PDFs and content you publish; their recipients may retain copies.

Graklink is based in the United States and uses providers with global infrastructure. Information may be processed outside your country. Applicable transfer protections depend on the service and jurisdiction; contact us for information about the safeguards relevant to your data. We do not promise that all data stays in one country.

Retention and your rights

We keep account and product records while needed to provide your services; transaction and tax records while needed for accounting, disputes and legal obligations; and security records while needed to prevent or investigate abuse. Support, feedback and communication records are retained to handle the request, maintain its history and resolve disputes. Availability requests remain until withdrawn or no longer needed for the requested update. These are purpose-based periods, not a promise that every record is deleted after the same number of days.

Retained aggregate history is different from detailed source observations, which may expire or be sampled. Report snapshots preserve the period and evidence used when generated. Account or connection deletion removes the applicable service data through the deletion process, subject to necessary legal records. Backup copies may persist until their recovery-retention period ends; a restoration must not reactivate a completed deletion.

Depending on the law that applies, you may request access, correction, deletion, a portable copy, restriction or objection to processing, withdrawal of consent, and review or appeal of a refused request. You can complain to your local data-protection authority. California rights can also include opting out of sale/sharing and limiting qualifying uses of sensitive information. We do not sell or share for behavioral advertising and do not penalize you for exercising applicable rights.

Send requests through Contact or support@graklink.com. Tell us what you need and the account or Graklink address involved. You do not need to buy a product to submit a privacy request. We may verify identity or an authorized agent's authority before disclosing or changing records. Provide only the information needed to locate the relevant records; do not send passwords, payment-card details or identity documents unless we explain why additional verification is necessary. If we refuse a request, you can reply asking us to review that decision. We respond within applicable legal deadlines and explain any permitted exception or extension.

Where applicable data-protection law requires a legal basis, processing necessary to provide a service you request relies on performing our contract with you or taking steps you request before a contract. Required accounting and responses to binding legal demands rely on legal obligations. Service security, fraud prevention and resolving support issues can rely on our legitimate interests in operating and protecting the service, subject to applicable limits and your rights. Optional processing that requires consent relies on that consent. These bases do not make every use necessary for a contract or remove a separate requirement for cookie consent. You can withdraw consent without affecting processing already lawfully completed. A customer using our measurement tools must separately assess its own responsibilities.

Optional email categories can be changed in dashboard Settings. Required account-security, purchase, billing and requested-support messages continue when applicable. Availability updates have a separate request-specific withdrawal link. Changing an email preference does not disable your product or dashboard reports.

Account deletion

Request deletion in Settings or through Support and complete the email verification. Confirmation disables your dynamic QR codes, short links and hosted content. The normal process keeps data for 14 days so you can use the emailed restoration link; immediate erasure skips that window and has no undo. Erasure requires completion of database, media and billing cleanup. If cleanup fails, it remains pending for retry rather than being reported as complete.

We remove account-associated personal information and uploaded media through that process. Minimal retired-address records prevent reassignment; required tax/payment records and security or legal records can be retained for their applicable purpose. The process cannot delete customer-exported copies, recipients' email or data a separate destination website holds.

Deletion does not automatically create a refund. Active subscriptions are canceled as part of the process, subject to completion of billing reconciliation. See Terms before confirming.

Children

Graklink's account service is not directed to children under 13, and we do not knowingly seek their personal information. If you believe a child has provided personal information, contact us so we can investigate and address it. Customers must not use our tools to collect children's information unlawfully.

Changes

We post revisions here with their update date. For a material change, we provide additional notice where required, such as an email or a notice in the service explaining the change and when it takes effect. If a new use of information requires your consent, we will ask before beginning that use. A policy update, an email delivery or continued use of the service does not by itself provide that consent.