Privacy

Who holds this data

Unbound Operators LLC, a Wyoming limited liability company, part of unboundoperators.com. We are the controller of the data described below, and the processors we use are listed further down.

What we collect

If you run a free scan

• The brand name, domain, category and competitors you type.
• Your email address, if you give one. It is optional.
• A one way hash of your IP address, used only to rate limit the free scan. We do not store the address itself.

If you create an account

• Your email address, for signing in.
• The brands, prompts, competitors and markets you configure.
• Your scans, results and spend records.
• If you subscribe, a Stripe customer identifier. Card details go directly to Stripe and never reach our servers.

What we do not collect

• No card numbers, ever.
• No passwords. Sign in is a one time link.
• No advertising trackers, and nothing that follows you to other sites.
• No IP addresses in readable form.

Who we send data to

Six processors, and only what each one needs:

• Supabase, which hosts the database and handles sign in.
• Vercel, which runs the application.
• DataForSEO, which relays our questions to the AI assistants and returns their answers. It receives the question text.
• Stripe, for payments, if you subscribe.
• Resend, to deliver email, if you give us an address.
• Google, for Analytics through Tag Manager, and only after you have accepted. Described in full under Cookies and analytics below.

Beyond those, the questions themselves reach the four AI assistants this product measures, because reaching them is the measurement: ChatGPT (OpenAI), Gemini (Google), Claude (Anthropic) and Perplexity. Two are reached through their consumer surface and two through their provider APIs, all of them via DataForSEO rather than by us directly. Each receives the question text and nothing else: no account identifier, no email address, and no record that connects the question to you.

We do not sell data, and we do not share it with anyone else.

How AI is used, and how it is not

This product exists to measure AI assistants, so it is worth being exact about which direction the data flows.

What goes to an assistant is a question about a category, built from the brand name, category and competitors you typed. "The best scheduling tools right now" is the shape of it. The question is asked from a clean session with no account history, which is the whole point: an answer coloured by our history would not be the answer a stranger gets. The assistants' replies are stored with the model that produced them and the date.

What does not happen is any model reading your data. No language model summarises, scores or makes a judgement about your account, and none is asked whether your brand looks good. Every figure on a report is counted from stored answer text by ordinary code. When the intake proposes a description of your site, it is parsed from that site's own HTML by a plain fetch, with no model and no provider involved.

Nothing you type is used to train any model by us. What the assistant providers do with a question put to them is governed by their own terms, and it is the same exposure as typing that question into the assistant yourself.

Data shared between customers

Three tables hold no customer data at all and are shared: the questions themselves, the answers engines gave, and the sources those answers cited. An answer belongs to the question, not to whoever is reading it, which is what lets one expensive measurement serve several customers in the same category.

Your brand, your competitors, your prompts and your results carry your account identifier and are enforced at the database level. That isolation is tested with two live accounts before any screen ships.

How long we keep it

• Free scan records and their reports: 12 months, so a link you forwarded keeps working.
• Account data: while your account exists, then 30 days.
• Rate limit hashes: 30 days.
• Billing records: as long as tax law requires, which is generally seven years.

Your rights

You can export everything we hold for your account at any time from the settings screen, as one file, without asking us. You can ask us to correct or delete your data by writing to the address below, and we will act within 30 days.

If you gave an email address to a free scan and want it removed, write to us with the domain you scanned and we will delete it.

Cookies and analytics

One cookie is always present: the one holding your sign in session. It is what keeps you signed in, so it has no alternative and no switch.

Everything else waits for you. We use Google Analytics, through Google Tag Manager, to see which pages people read and how they arrived. Nothing from Google is loaded, and no analytics cookie is written, until you have said yes. That is stricter than the usual banner, which typically sets its cookies first and asks afterwards. Decline and no request is made to Google at all, which means our numbers under-count the people who decline. We would rather have a figure that is honestly low than one that counts people who said no.

If you accept, Google Analytics sets its own cookies to tell one visit from the next, and receives the page you viewed, roughly where you are and what you are browsing with. It does not receive your name, your email or anything you type into the product. Google is the controller of that data under its own terms.

You can change your mind at any time. Clearing this site's stored data in your browser removes the record of your choice, and the question comes back on your next visit. Declining then is enough to stop it; there is no account setting to hunt for.

Contact

Unbound Operators LLC, Wyoming, United States.
hello@withunbound.com
unboundoperators.com

Data is stored and processed in the United States. If you are in the EU or UK, transfers rely on the standard contractual clauses our processors publish, and you should have counsel confirm that wording before we take EU customers.