1. Who we are and our role
STORYSTEPPER LIMITED, Room 701, Unit 108, 7/F, Tower B, New Mandarin Plaza, 14 Science Museum Road, Tsim Sha Tsui, Kowloon, Hong Kong, operates VozRadar.
For your account, billing, security logs, our website and our own analytics, we act as the controller: we decide why and how that data is processed.
For the content you put into a workspace and the data you connect to it (your website's pages, tracked prompts, competitor names, Search Console and analytics data, repository file paths, report recipients, alert destinations), you act as the controller and we act as your processor. We process it only to provide the service to you and on your instructions. If your organisation needs a data processing agreement, contact [email protected].
2. What we collect
Account and sign-in
- Email address, display name, and a password hash if you sign up with email and password. Passwords are never stored in clear.
- If you sign in with Google: your Google account identifier, email address and profile name, requested with the
openid email profilescopes. - Six-digit one-time codes sent by email at sign-up, for password reset and when you change your email address. Codes are stored hashed, expire, and are single-use.
- Session records: a hashed session token, the browser's user agent, the IP address at sign-in, and creation, last-seen and expiry times. You can see and revoke your active sessions under Account settings.
- Failed sign-in attempts (email and IP address) used to throttle password guessing.
- The result of a Cloudflare Turnstile challenge on sign-up, sign-in and password-reset forms, used to block automated abuse.
Workspace and billing
- Workspace name, your role in it, members you invite (their email address) and invitations you send.
- If you subscribe to a paid plan: the Stripe customer and subscription identifiers linked to your workspace, plan, status and renewal dates. Card and billing details are entered directly on Stripe's checkout page and are not stored by us.
- A record of connectors added, configured, synced or removed, and by whom.
Content you create in a workspace
- Projects: the domain you connect, brand name, industry, competitor names and aliases, markets and cities.
- Tracked prompts and prompt variants, which are benchmark questions you approve, not private conversations.
- Public pages crawled from the domain you connect and from competitor domains (public pages only).
- Approved brand facts, recommendations, experiments and notes.
- Client reports, including a share link if you create one, and how many times it was opened.
- Alert channels: email addresses, Slack incoming-webhook URLs or webhook URLs you configure. Targets are stored encrypted.
Data we generate
- AI answers collected from the answer surfaces you enable, with their citations, model, surface, locale and location, and the raw provider response.
- Classifications of those answers (whether a brand is mentioned, recommended, cautioned against), extracted claims and fact-check verdicts.
- Aggregated visibility metrics, source statistics and demand estimates.
- Provider usage and estimated cost per execution, used for plan limits and budgets.
Optional connected data
Only if you connect them, and only with read-only access:
- Google Search Console: query and page performance for your property in the project's market, used as a demand signal for prompts.
- Google Analytics 4: sessions referred by AI assistants, per day and landing page, stored as aggregates. No user-level analytics data is imported.
- GitHub: the list of file paths in the repository you choose and the metadata of pull requests you link to experiments. VozRadar reads file paths and PR metadata, never file contents. Repository data is not sent to AI providers.
Connector tokens are stored encrypted and can be disconnected at any time from the project's integrations page, which stops further syncing.
Website visitors
Our edge provider, Cloudflare, sees the IP address and request metadata of every request in order to serve and protect the site.
If you join the prelaunch waitlist, we store your email address, where on the site you joined, the date, and your IP address for abuse prevention. We send one confirmation email and contact you when access opens. Cloudflare Turnstile checks the form for automated abuse.
Analytics and error tracking
We use PostHog, a product-analytics service, to understand how our website and app are used and to find errors. It runs on our public pages, the sign-in and sign-up pages, the signed-in app and shared report links. Your browser sends analytics to our own domain and our server forwards it to PostHog, so PostHog receives your IP address and request metadata through us. We collect:
- the pages you view, the page you came from and any campaign (UTM) parameters in the link;
- clicks on page elements, including which call to action you clicked;
- browser, device, operating system and screen size;
- an approximate location (country, region and city) derived from your IP address;
- page performance measurements (web vitals);
- JavaScript errors and failed requests to our API: the error message, stack trace and page URL;
- product events such as joining the waitlist, signing up, signing in and out, starting Google sign-in, creating a project, completing onboarding steps, creating a report, accepting an invitation, starting or completing a checkout, changing plan, opening the billing portal and deleting an account.
Session recording is not enabled. We do not use analytics for advertising, we do not sell it, and we do not track you across other sites. Whether analytics may store anything in your browser, and whether it is linked to your account, depends on your answer to the cookie banner, described in section 7.
3. Why we process it and on what legal basis
| Purpose | Data | Legal basis (GDPR) |
|---|---|---|
| Creating and securing your account, sessions, one-time codes, sign-in throttling | Account and sign-in data | Contract; legitimate interest in security |
| Running the prelaunch waitlist and preventing form abuse | Waitlist email, source, date and IP address | Your request to join; legitimate interest in abuse prevention |
| Analytics and error tracking without storing anything on your device (cookieless mode, section 7) | Analytics data listed in section 2; a daily-changing hash of IP address and user agent | Legitimate interest in measuring and improving our website and app |
| Analytics with cookies and browser storage, and linking analytics to your account when you are signed in | Analytics data listed in section 2; analytics cookie identifier; user ID, email address and display name | Consent, given in the cookie banner and withdrawable under Cookie settings |
| Providing the service: crawling, prompt discovery, collecting and classifying AI answers, recommendations, reports, alerts | Workspace content, generated data, connected data | Contract (as processor, on your instructions) |
| Billing and plan enforcement | Workspace and billing data, usage metering | Contract; legal obligation (tax and accounting) |
| Transactional email: codes, invitations, alerts you configured | Email address, message content | Contract |
| Abuse prevention, budget limits, incident response | Sign-in logs, usage metering, audit records | Legitimate interest in running a reliable service |
| Responding to requests and legal obligations | Contact details, records | Legal obligation; legitimate interest |
We do not sell personal data, and we do not use your workspace content to train models.
4. Who processes data on our behalf
We use the following subprocessors. Each receives only what its purpose requires.
| Subprocessor | Purpose | Data categories |
|---|---|---|
| Cloudflare, Inc. | Edge proxy and TLS for the website and app; Turnstile bot protection on account forms; R2 object storage for raw AI-answer responses and database backups | IP address and request metadata; Turnstile challenge result; raw provider responses; encrypted backups |
| Stripe, LLC and its affiliates, including Link (merchant of record through Stripe Managed Payments) | Checkout, subscriptions, receipts and invoices, taxes, refunds, payment disputes, subscription management at link.com | Name, email, billing address, tax ID and payment details you enter at checkout; workspace identifier |
| Resend, Inc. | Transactional email | Recipient email address; email content (one-time codes, invitations, alert messages) |
| PostHog, Inc. (PostHog Cloud, US region) | Product and website analytics and error tracking for our own site and app | Analytics data listed in section 2; IP address and request metadata; if you accept cookies: an anonymous analytics identifier and, when signed in, your user ID, email address and display name |
| Google LLC | Sign in with Google; Search Console and GA4 connectors (read-only, optional) | Google account id, email, profile name; Search Console query/page performance; aggregated GA4 AI-referral sessions |
| GitHub, Inc. | Optional read-only GitHub App | Repository file paths, pull-request metadata; installation tokens |
| OpenAI | Answer surface (web-grounded API) | Prompt text, market, language, city or region on the project |
| Google LLC (Gemini API) | Answer surface (Google Search grounding) | Prompt text, language |
| Perplexity AI, Inc. | Answer surface (web-search API) | Prompt text, country, city or region on the project |
| Anthropic, PBC (Claude API) | Answer surface (Claude with web search), only when a workspace on a plan that includes it selects it | Prompt text, country, city or region on the project |
| SerpApi, LLC | Google AI Mode and AI Overview search-result capture; Google Trends lookups for demand estimates | Prompt or keyword text, location, language |
| Anthropic, PBC | Classifying collected AI answers; extracting a brand profile from your public pages; generating and localizing candidate prompts; content briefs; fact checks against your approved brand facts | AI answer text, text of public pages crawled from your site, brand facts, prompt text, competitor names |
| Voyage AI | Semantic de-duplication of candidate prompts (only where enabled) | Prompt text |
| OVHcloud (OVH US, United States) | Servers and PostgreSQL database that run the service | All service data |
As merchant of record, Stripe also uses the details you enter at checkout for its own purposes, such as fraud prevention, tax and its legal obligations, under the Link Privacy Policy at link.com/privacy.
Prompts sent to answer surfaces are your approved benchmark questions plus the market and location set on the project. They do not include your name, email or account details.
Alert destinations you configure (Slack workspaces, webhook endpoints, email recipients) receive the alert content you asked for. They are your recipients, not our subprocessors, and are governed by your agreements with them.
Infrastructure alerts about the service itself (aggregate system metrics, no customer content) may be delivered to our operators through Telegram.
We will update this table when a subprocessor changes. Material changes are announced under section 12.
5. International transfers
Several subprocessors above are established in the United States or process data there; analytics data sent to PostHog is processed and stored in the United States. Where data leaves the European Economic Area, the United Kingdom or another jurisdiction with transfer rules, we rely on the recipient's Standard Contractual Clauses or an adequacy decision, plus the encryption and access controls described in section 8. Our own servers are located at OVHcloud (OVH US, United States).
6. How long we keep it
| Data | Retention |
|---|---|
| Account, workspaces, projects, prompts, answers, classifications, reports, experiments | For as long as your account or workspace exists |
| Waitlist email and signup date | Until launch, or until you ask us to remove it |
| Waitlist signup IP | Removed from the waitlist record after 30 days |
| Sessions | 30 days of inactivity, then removed; you can revoke any session earlier |
| One-time email codes | Deleted about a day after they expire |
| Failed sign-in records | Deleted after one day |
| Raw provider responses (object storage) | 90 days after the answer was collected; the answer text and citations stay with the project |
| Connector tokens and imported connector data | Until you disconnect the connector or delete the project |
| Report share links | Until they expire or you turn them off |
| Analytics events | Kept in PostHog; deleted on request to [email protected] |
| Billing records | As required by tax and accounting law |
| Database backups | 30 days |
Deleting your account. Under Account settings → Delete account, you delete your account and every workspace where you are the only member, with all of its projects, answers and reports. Raw provider responses stored for those workspaces are removed from object storage shortly afterwards, normally within the hour, and copies in database backups expire within 30 days. In shared workspaces your membership is removed and records you created (experiments, monitoring batches) are kept for the other members without your identity attached. If you are the only owner of a workspace that other people use, you must hand over ownership or remove them first. Deleting your account also resets the analytics identity in your browser.
8. Security
- Traffic is encrypted in transit; the origin is reachable only through Cloudflare.
- Passwords are hashed. Session tokens and one-time codes are stored only as hashes.
- Connector credentials and alert targets are encrypted at rest with a key held outside the database.
- Workspace membership is checked server-side on every request. Automated tests guard against cross-workspace access.
- Raw AI-answer responses live in private object storage with no public URLs.
- Provider API keys stay on the server and are never sent to the browser.
- Metrics that carry usage and spend are served on an internal port that is never exposed.
- Crawled web pages are treated as untrusted input; instructions found in page text are never executed.
- Nightly encrypted database backups.
No system is perfectly secure. If we learn of a breach affecting your personal data we will notify you without undue delay as the law requires.
9. Your rights
Depending on where you live, you may have the right to access, correct, export, restrict or delete your personal data, to object to processing based on legitimate interest, and to withdraw consent where consent is the basis. You also have the right to complain to your supervisory authority.
Much of this is available in-product: edit your profile and email address, revoke sessions, export observations and citations as CSV from a project, disconnect connectors, change your analytics cookie choice under Cookie settings in the site footer, and delete your account. For anything else, write to [email protected]. We answer within one month.
If you are a member of a customer's workspace and the request concerns data that customer controls, we may refer you to them or act on their instructions.
10. Children
VozRadar is a business tool and is not directed at children under 16. We do not knowingly collect their data.
11. Public web data and our crawler
To measure visibility we crawl public pages of the domain you connect and of competitor domains. The crawler identifies itself with the VozRadarBot user agent, follows robots.txt, rate-limits each host and does not attempt to bypass logins or paywalls. Site owners who want to limit crawling can do so in robots.txt. AI answers we collect may name people or organisations that are public in that context; we use them only to compute visibility and source metrics for the project that asked the question.
What the crawler visits, what it stores and how to block or slow it down are described on our crawler page.
12. Changes to this policy
We will post changes here and update the effective date. For material changes we will notify account holders by email or in the app before they take effect.
13. Contact
Privacy questions and requests: [email protected]
Postal: STORYSTEPPER LIMITED, Room 701, Unit 108, 7/F, Tower B, New Mandarin Plaza, 14 Science Museum Road, Tsim Sha Tsui, Kowloon, Hong Kong
Company registration number: 80655938-000-06-26-8