On this page 14
- Data controller
- What data we process
- What we use it for and on what basis
- Data we collect from public sources
- Optional website analysis
- The console assistant
- Google Search Console
- Who else is involved
- Transfers outside the European Economic Area
- How long we keep it
- Your rights
- Cookies and audience measurement
- Automated decisions
- Security and changes
Data controller
The controller of the data you provide on this website and in the client workspace is:
- Operator
- Manuel Sampedro Pérez
- Trade name
- AnswerGravity
- Address
- Calle de Serrano, 61, 28006 Madrid, Spain
- Tax ID
- 49573698G
What data we process
What you give us when you ask for a review, a diagnosis or a proposal, when you create or manage an account and when you write to us: name, work email, company, website, market and the message you choose to send.
If you use the client workspace, also your workspace data: questions, model answers, cited sources, public conversations found by listening, reports, drafts, activity records and technical connection and session data, such as the IP address we use to limit requests and protect sign-in.
If you sign in or create your account with Google or Apple, they give us your name, your email and that account’s identifier, and we keep them to recognise you when you come back. We do not receive your Google or Apple password or any other data from those accounts. If you choose to hide your email with Apple, we receive an Apple address that forwards our emails to you.
If you buy, your billing and payment data: name or company name, billing address, tax or VAT number, email, plan, amounts, taxes, payment status and invoices. If you start a free trial, Stripe also asks for a phone number. Your card details are handled by Stripe; AnswerGravity never sees or stores them.
Please leave out sensitive data and any personal data about others that is not needed.
What we use it for and on what basis
Answering your request and preparing a proposal: pre-contractual steps you ask for.
Providing the service you bought, authenticating your account, managing the subscription, payments and invoices, sending operational messages such as the reminder before a free trial ends, and keeping the evidence of measurements: performance of a contract.
Protecting the service, preventing fraud, offering one free trial per person and improving reliability: our legitimate interest.
Showing each client the public conversations about its category: the client’s legitimate interest and ours, as the next section explains.
Measuring how the site and the console are used and which steps lead to a purchase, as explained under “Cookies and audience measurement”: with PostHog, your consent in the European Economic Area, the United Kingdom and Switzerland and our legitimate interest elsewhere; with our own count and Cloudflare Web Analytics, our legitimate interest.
Issuing and keeping invoices and meeting other tax and legal obligations: legal obligation.
We do not send marketing without a valid basis, and you can refuse it at any time.
Data we collect from public sources
For measurements and listening, Bright Data collects on our behalf AI assistant answers and public search results, including forum threads on sites such as Reddit and Quora, for the questions and keywords each client chooses. For each thread we keep the title, the link, a text snippet, the community and its search position. This content is public and may include the username of whoever posted it or other details that person made public.
We use it only to show the client what is said in public about its category and where it makes sense to take part, based on the client’s legitimate interest and ours. We do not build profiles of people and we do not contact the authors. It is kept while the client keeps that brand in its workspace, and it is deleted with the brand or the workspace.
If you appear in that content, you can object to this processing or ask us to erase it by emailing hola@answergravity.com. Include the link to the content so we can find it.
Optional website analysis
When you ask us to analyse a website or discover competitors, we read its homepage and up to six linked public pages. We send Anthropic (or OpenAI, if Anthropic is unavailable) excerpts from those pages, the business name and the names and domains of competitors you already track to prepare a profile and sourced suggestions. We do not send your saved questions, Search Console data, reports or access credentials in this analysis. You can review, change or remove every suggestion before saving. We reuse the result within your workspace for 24 hours to avoid repeating the analysis. It is not a visibility measurement and does not guarantee that an AI will recommend your business.
The console assistant
When you write to the console assistant, we send Anthropic your message, the earlier messages of that conversation and the workspace data it looks up to answer, such as your questions, measurements, competitors, cited sources and listening conversations. We do not send it payment or access data. Answers are generated by AI and may contain mistakes: check them before acting on them. We keep conversations in your workspace so you can come back to them, and they are deleted with the workspace.
Google Search Console
When this integration becomes available, connecting it will be optional. If a workspace owner or admin connects it, Google will ask for read-only permission. We read the properties that Google account can access and, for the property you choose, searches from the last 28 days with final data, their impressions, clicks, click-through rate and average position. We cannot change your website or its Search Console data. Workspace members with access to every brand and without a read-only role can view this information.
We use this data to suggest questions you can review, edit and import. Suggestions are prepared in your browser without sending them to an AI model. We do not store the list of searches or their metrics in our database: only the questions you choose to add. Once you add them to tracking, those questions are sent to the selected measurement services and assistants to obtain their answers, like other questions in your workspace.
We store the permission that keeps the connection active in encrypted form, together with who connected it and when. You can disconnect it in Settings, “API & MCP”: we delete the stored permission and ask Google to revoke access. You can also remove it from your Google Account connections. Disconnecting does not delete questions you have already imported; you can manage them in Prompts or request erasure as described under “Your rights”.
The connection relies on your authorization and is used only for the feature you request. We do not sell this data or use it for advertising. The use and transfer of information received from Google APIs adhere to the Google API Services User Data Policy, including its Limited Use requirements.
Who else is involved
These providers process data on our behalf, as processors, and only to provide their service to us. Stripe also processes some data as an independent controller, to prevent payment fraud and meet its own legal obligations, under its privacy policy. We do not sell personal data or share it with third parties for their own purposes, unless the law requires it.
- Railway
- Hosting of the website, the application and its background processes.
- Cloudflare
- Hosting the collector and storing the aggregate counters, and the traffic statistics of Cloudflare Web Analytics, described under Cookies and audience measurement.
- Neon
- The application’s database, hosted in the European Union (Frankfurt).
- PostHog
- Product analytics for the site and the console, with the data hosted in the European Union (Frankfurt), described under Cookies and audience measurement.
- Resend
- Sending the application’s emails, such as verification codes, notices and reminders.
- Spaceship (Spacemail)
- Email for the answergravity.com domain: it receives the messages you send us.
- Google (Gmail)
- The operator’s mailbox, which receives notices of new requests and replies to our emails.
- Google (Calendar y Meet)
- Booking the 15-minute call: it receives your name, your email and the time you choose, and sends the invitation with the Google Meet link.
- Google and Apple
- Signing in with your Google or Apple account, if you choose to. Each provides it as a controller, under its own privacy policy.
- Stripe
- Card payments, subscriptions, tax calculation, invoices, the customer portal and fraud prevention.
- Bright Data
- Collection of AI assistant answers and public search and forum results for measurements and listening.
- Anthropic
- Claude models: analysing public pages to prepare a business profile and suggest competitors and questions, preparing text drafts, the console assistant’s answers and, when active, measuring Claude’s answers.
- OpenAI
- The same analysis and drafting features as a fallback, when Claude is unavailable or its spending limit is reached.
Transfers outside the European Economic Area
Neon keeps the database, and PostHog the analytics data, in the European Union. Railway, Resend, Spaceship, Google, Apple, Stripe, Cloudflare, PostHog, Anthropic and OpenAI are US companies or part of groups based there, and Bright Data is based in Israel, so they may process data outside the European Economic Area. When they do, the safeguards required by the GDPR apply: an adequacy decision of the European Commission, such as the one for Israel or the EU-US Data Privacy Framework for companies certified under it, or standard contractual clauses approved by the Commission. You can ask us which safeguard applies to each provider.
How long we keep it
Requests, for as long as needed to answer you and follow up on the proposal.
Account data, while the account is active and afterwards for as long as needed for claims, security or legal obligations. By default, measurement evidence is kept for 365 days and reports and audit records for 730 days, unless the workspace sets another permitted period or a legal hold applies. Sessions expire after 30 days.
Invoices and billing data, for the periods Spanish tax and commercial law requires, generally four to six years.
PostHog analytics data, as explained under “Cookies and audience measurement”.
Free trial fingerprints, twelve months.
Your rights
You can ask for access, rectification, erasure, restriction, objection and portability, and withdraw consent without affecting earlier processing, by emailing hola@answergravity.com. We will ask for reasonable information to verify your identity before answering.
If you disagree with our answer, you can complain to the Spanish Data Protection Agency.
Automated decisions
AI features prepare evidence or drafts that a person reviews. We make no automated decisions with legal or similarly significant effects on anyone.
If you start a free trial, we keep for twelve months a fingerprint of your card, email, phone and connection address, computed with a secret key, to check automatically that you have not had another trial. The fingerprint does not contain that data in readable form and we use it for nothing else. If the check leaves you without a trial, you can still subscribe or write to us to review it. Stripe may also use automated systems to detect fraudulent payments.
Security and changes
We use workspace access controls, securely stored passwords, protected session cookies, defences against forged requests, request limits, encryption in transit and audit records. No system removes every risk. If we change anything material, we will publish it here with a new effective date.