Feder Privacy Policy
Effective date: 2026-10-16
Last updated: 2026-11-19
1. Who this is for
Feder is operated from the United States. Your information is stored and processed in the US. If you use Feder from outside the US, you understand that your information is transferred to the US and handled as this policy describes. This policy describes what we collect, why, and what we do with it.
Feder is a business product for businesses and the people who run them. It is not intended for anyone under 18, and we do not knowingly collect data from children.
2. What we collect
Account information. When you sign up, we collect your name, work email, business name, the territory you choose, and login details. We also keep the time zone your browser reports, so dates show in your time. If you invite staff or client viewers, we collect their names and emails too. If you add an email address as an alert destination, we keep that address so we can send alerts to it. If you came to Feder from a site reading on getfeder.com, we keep the address of the site you read so the app can name it.
Signing in. You sign in with a one-time link or code that we email you, or with your Google account. Feder has no passwords for members. The one exception is Feder's own account for app directory reviewers, whose password is stored only as a scrambled (hashed) copy. If you add a passkey or an authenticator app, we store only what is needed to check it. Where it is offered, you can also sign in with GitHub; we then receive your GitHub id, user name, name and verified email. Your browser keeps a sign-in cookie; we store only a scrambled (hashed) copy of it, so the stored copy cannot be used to sign in. We do not store your IP address, browser or location with your sign-in. To limit repeated sign-in attempts, we keep a scrambled (hashed) copy of the IP address and email that asked for a sign-in email.
Google connections. If you sign in with Google, we receive your name and email address. You can also connect these Google products to Feder through Google OAuth: Google Analytics (GA4), Search Console, YouTube, Tag Manager, and Google Ads. From each one, Feder reads only what it needs to check that your tracking is working. If you ask Feder to fix a GA4 key event, Feder asks Google for one-time edit access, makes only that change, and then asks Google to revoke that access. That edit access is never stored. We store the OAuth tokens needed to keep your connections working. Tokens are stored encrypted and are used only to check those sources on your behalf.
Google user data. Feder's use and transfer to any other app of information received from Google APIs will adhere to the Google API Services User Data Policy, including the Limited Use requirements.
YouTube. Feder uses YouTube API Services. When you connect YouTube, Feder reads your channel identifier and title, lifetime view and subscriber counts, and identifiers, titles and view counts for up to ten recent videos. Feder may also read the authorizing Google account's email address to label the connection. Feder stores the channel sample and connection/check records to show your readings and whether the connection works. Nango manages this connection and its credentials for Feder; Google and Nango process the requests needed to provide it. See the Google Privacy Policy.
You can disconnect YouTube in Feder. In addition to the deletion controls and requests in section 7, you can revoke Feder's access through Google's security settings. Revoking access and deleting stored data are separate actions. For deletion requests or questions or complaints about YouTube data and our privacy practices, contact support@getfeder.com.
Connected services. You can also connect other business tools to Feder, for example customer tools, website platforms, email tools, and scheduling tools. You connect in one of two ways: by signing in to that service and approving Feder's access, or by giving Feder an API key or token from your own account with that service. Either way, the access is read-oriented: Feder reads what it needs to check whether the source is working. A few services, such as Google Ads and Microsoft Advertising, only offer access that would also allow changes; Feder only reads with it. We store the sign-in tokens or API keys needed to keep those connections working, or, for apps connected through Nango, Nango stores them for us. They are stored encrypted and are used only to check those sources on your behalf. You can disconnect a service at any time, and disconnecting stops our access.
Connected-source checks. For each connected service, we keep the identifiers, check results, and configuration facts needed to tell whether it is working (for example, whether expected data is arriving). From GA4 we store daily counts by event name. We do not store individual GA4 hits. For some connected apps, Feder also keeps one number per day for a signal, such as visits, sign-ups, orders, bookings, or money taken. It keeps only that daily total, never the records, customer details, or text behind it.
Connect with code. When a site uses Feder's page tag, we store each accepted page event so we can tell whether tracking is arriving. Each stored event has the event name, the time our server received it, the time the browser claimed (kept for debugging), whether its parameters were over our size limit, and whether the page's origin was one the site allows (until a site's address is confirmed, every origin counts as allowed). The tag sends a small set of parameters with each event: the automatic page view sends the page path (the address path only, with no query string, cut off at 200 characters) and, when the browser has one, the referring site's hostname. We read those parameters when the event arrives and do not store them. Feder does not add a name, email, or visitor id of its own. If the tag is on the page more than once, we note that for the day. The tag sets no cookie, uses no local storage, and adds no visitor id that follows a person from visit to visit. We do not save an IP address with the event. A custom event can send up to 20 flat text, number, or yes/no parameters. Like the page view's, we read them when the event arrives and do not store them. The tag asks the site not to put names, emails, or free text in those parameters. We also keep, per site and per UTC day, the count of accepted events, the count of events we did not store because the day was full, and the count of events by page origin. The origin is the site address (such as https://example.com), not a page path. Free stores up to 5,000 events per site per UTC day. Circle, including Lifetime, stores up to 25,000. Stored events are deleted by the site's regular check once they are older than the history the check reads plus 7 days of margin: about 7 weeks on Circle, whose check runs daily, and up to about 9 weeks on Free, whose check runs weekly. If a check cannot run, deletion waits for the next one.
Alert and incident history. When Feder detects a tracking issue, we store the incident record: what was detected, when, and which alerts were sent. This is the core of the service. Feder also keeps an internal copy of each tracking incident (the GA4 property id, the event names, and the expected and actual daily counts) to improve how it detects problems. That copy is removed when the incident is removed.
Billing information. Payments are processed by Stripe. We do not see or store your full credit card number. We store plan, billing status, and transaction records needed to run your subscription.
Support communications. If you email support@getfeder.com or send a note from the in-app help, we keep the correspondence so we can help you and improve the service. With a note we also record the page you were on and your screen size, and any screen picture you choose to attach; these go to our support inbox by email. A key or token in a note is removed before the note is emailed to our support inbox or kept. For questions you ask the in-app help, we keep the question, the answer, the page you were on and whether you said it helped. If you write to us from the help on a public page without signing in, we send your email, your message and the page you were on to our support inbox.
Slack channel reading (optional). If an Admin connects Slack channel reading, Feder reads messages in the public channels you select, and only those. It does not read direct messages, private channels, shared channels, files or search. Feder keeps the workspace and channel names and ids and who connected it. It counts each message as it arrives and does not store the message text. It keeps daily totals per channel (messages, replies and bot posts) for up to 90 days while reading is on. Results show channels and time windows, never a person or a message.
AI assistants you connect (optional). You can connect an AI assistant, such as ChatGPT, Claude, Cursor or Grok, or Feder's Zapier integration, so it can read your Feder account. Everything in this paragraph about an assistant applies to Zapier too. You sign in to Feder and approve read-only access for that one assistant. Feder keeps the assistant's name and, where it gives one, its web address, which person approved it and when, when it last read, and only scrambled (hashed) copies of the codes and tokens it uses. To limit how many assistants can register, Feder also keeps a scrambled (hashed) copy of the network address that registered one. Each time you ask it, the assistant reads what Feder's read-only tools return about the account you approved: its businesses, monitored sites, the events each site watches, open tracking breaks, breaks fixed in the last 30 days, and whether each check is on and when it last succeeded. It cannot start a check, change a setting, send anything or see another account. What the assistant's provider does with what it reads is covered by that provider's own privacy policy. You can disconnect an assistant at any time in Settings, AI assistants, which ends its access straight away, and access an assistant has not used for 30 days ends by itself. If you create an access key instead, we store only a scrambled (hashed) copy of the key and when it was last used, and the assistant can read your account data until you revoke the key.
Invitations, codes and referrals. If you invite someone to a free year or as a founding member, we keep their name and email to send and track the invitation. When a code, invitation or referral link is used, we keep a record of it so it cannot be used again.
Scan results and waitlist. If you ask for scan results or join the waitlist, we keep your email, the site you scanned and the campaign tags in the link you came from. Where used, joining the waitlist also adds your email to our email list at Resend (section 5) so we can send you the waitlist emails. If Feder has written to you about Feder, we keep your name, email and business so we do not write twice.
Stopping Feder's outreach. Feder sends some of its own outreach emails to businesses through Instantly. Where this is used, when you create an account Feder sends your sign-in email to Instantly so that outreach stops writing to you: Instantly adds the email to its do-not-contact list, and any outreach contact with the same email is marked as a customer or removed.
Shared patterns (optional). If you join shared patterns, Feder groups your account with others in the same territory so a common tracking break is easier to recognize. No person or business is named. You can leave any time in Settings, Privacy & data, and leaving never changes what Feder watches on your own sites.
Share. Some pages let you download an image or copy a short summary of your results. Feder makes these on your device from what the page already shows. Feder does not upload or publish them and makes no link to them. What you share, and with whom, is up to you.
Product usage counts (optional). If your product usage counts are on, we keep one count each time your account views a part of Feder or reaches a milestone (for example finishing setup or starting a plan): the feature and the time, nothing else. Counts start on for accounts in the United States and off everywhere else, and you can turn them on or off any time in Settings, Privacy & data. When counts are turned on, we also count the milestones your account already reached (for example signing up, finishing setup or its first check) at the time they happened. Turning them off deletes the counts already kept. We use them only in totals across many accounts, never to single one out, and counts from Feder's own accounts are left out. They are kept while your account exists.
Journey analytics (optional). Journey analytics is an optional choice, separate from product usage counts: allowing it records selected page visits and product activity as events tied to a random id for your browser, and to your account while you are signed in, for up to 7 days, while identifier-free page-open counts (route category and a coarse time bucket, plus viewport and referrer categories only when allowed) are kept for up to 30 days either way. You can also allow an occupancy grid: coarse pointer or tap cells, scroll milestones, and foreground time. That grid stores no text, form contents, or keystrokes, and those counts are kept for under 7 days. If you allow content experiments, Feder may show you one of two versions of a page, chosen at random, and records which one you saw; while a scan-results trial is offered, Feder can also measure those trial signups. Experiments do not change prices, billing, or consent. If you allow journey analytics, Feder also keeps on your account where you first and last arrived from (the campaign tags of those links) and which experiment versions you saw, up to five. That record stays while your account exists. A Lifetime membership does not include a scan-results trial. When checkout and billing events are offered, a fresh choice also covers subscription start, cancel, status, and scheduling events after you open checkout or billing. The account link is kept for up to 7 days, and events contain no payment details. A Lifetime membership has no checkout or billing events.
3. How we use it
We use your data to:
- Provide monitoring, detection, and alerts.
- Manage your account, plan, and billing.
- Send service emails (alerts, billing notices, security notices, and account setup and progress emails such as the welcome).
- Send alerts to the places you choose, such as a Slack channel or webhook you add.
- Respond to support requests.
- Improve reliability and fix bugs.
We use AI (Anthropic's Claude) to answer the questions you ask in the in-app help. When you ask, your question, the page you are on, and a summary of your own account (for example your businesses, connections, checks, and open issues, and on Circle your readings, such as signal values, money totals and daily totals from connected apps) go to Anthropic to write the answer. A key or token you paste into a question is removed first. Feder recognizes common key formats, not every one, so please do not paste keys or passwords. Feder also uses Claude to spot common problems across recent help questions and notes from all accounts so we can fix them. Before they are sent, Feder removes keys, emails, links, common web addresses (such as .com addresses) and long numbers, and the summaries Feder keeps never name your business. This is separate from shared patterns. Your data is never sold or used to train AI models.
We do not use your data for advertising, and we do not sell your personal information.
4. Cookies
This section covers the Feder app. Our website, getfeder.com, has its own privacy notice at getfeder.com/privacy, which covers the website's cookies and analytics.
In the Feder app:
- Strictly necessary: login sessions, security, load balancing. The app cannot work without these.
- Feder's own analytics: your analytics choice is kept in your browser's storage for up to 7 days. If you allow journey analytics, Feder sets its own cookie holding a random id for up to 7 days, and your browser's storage keeps that id and where your visit came from. If you decline, only your choice is kept. A Feder team member's opt-out cookie lasts 30 days.
- Preferences: remembers choices like display settings.
- On this device: after you sign in, the app keeps a small greeting for the sign-in page in your browser's storage: your first name, your Roots score, and which way money went, never an amount. It is never sent to Feder, it expires after 30 days, and signing out removes it.
- Google Tag Manager and Google Analytics: the app loads Google Tag Manager. Analytics and advertising storage start off, so Google Analytics sets no cookies unless you allow journey analytics, and Feder never turns advertising storage on. Like any web request, loading it can tell Google your IP address and browser type.
5. Who we share data with (subprocessors)
We share data only as needed to run the service:
- Stripe - payment processing and subscription billing.
- Google - sign-in, and access to the Google products you connect (section 2), at your direction. Google Tag Manager and Google Analytics in the app (section 4). Where used, Feder also sends your site's address to Google's Chrome UX Report to read your site's public page-speed data.
- Nango - managing connected-service authentication and connection data, including holding the sign-in tokens and keys for apps connected through it.
- Inngest - scheduling and running background jobs, including monitoring checks.
- Resend - sending service emails, such as alerts and invitations, and support notes to our inbox. Where used, Resend also keeps the list of who gets which service email: your email, your first name, your business name, your account's status, whether you have connected a source, and which app you connected first.
- Slack - delivering alerts to the channels you choose and, if you connect it, channel reading (section 2).
- GitHub - sign-in, where offered.
- Instantly - sending Feder's own outreach emails to businesses. Where used, it receives a new member's email so that outreach stops (section 2).
- The services you connect - for example Microsoft Advertising, Meta, TikTok, HubSpot, Salesforce, Square, Stripe, Calendly or Braintree receive Feder's requests to read your own account there, at your direction (section 2).
- Loom - hosts a video on some invitation pages; playing it loads it from Loom.
- Anthropic - AI processing for answers in the in-app help and for service issue summaries. Anthropic does not train its models on this data.
- Hosting - application and database hosting: Vercel and Neon.
We do not sell personal information to anyone. We do not share data with advertisers or data brokers.
We may disclose data if required by law (for example, a valid court order), and we will notify you unless legally prohibited.
6. Data retention
- Cancelling a paid plan does not delete anything: your account keeps being watched on the Free plan. Free keeps 90 days of history. Circle keeps 24 months of history. That is how far back you can look. Feder does not yet delete older history on a schedule; it is kept while your account exists. When you delete your account, your workspace data is deleted, and backups containing it age out within 90 days. Records of the codes, invitations and referrals your account used are kept so they cannot be used again. Where used, the do-not-contact entry made at Instantly when you signed up (section 2) also stays, so Feder's outreach does not write to you again.
- When you disconnect a Nango-managed connection or request deletion, Nango soft deletes the connection data immediately and hard deletes it within 31 days.
- Nango also keeps a log of each request Feder makes through it (the address, method, result and headers, with credentials removed, never the data returned) for 15 days under its standard terms. Disconnecting does not remove these logs; they age out on their own.
- Questions you ask in the in-app help, and notes you send from it, are kept in Feder for 90 days, then deleted. Notes you send to support also arrive by email and are kept as support records.
- Anthropic deletes what Feder sends it for in-app help answers and issue summaries within 30 days under its standard commercial terms, unless the law requires longer or its safety systems flag a request, in which case it may keep that request for up to 2 years.
- Inngest, which runs Feder's background jobs, keeps a record of each job (internal ids and results, not your account content) for a limited time that depends on its plan, 14 days or less on its standard plans.
- Apart from the incident copy in section 2, we do not keep separate long-lived copies of your data. Totals that identify no one, such as daily counts by event name, may be kept longer.
- Billing records needed for tax and accounting may be kept longer as required by law.
7. Your privacy requests
You can download a copy of your main account data yourself: open Settings, Privacy & data and choose Download my data. It includes your business, its people, connections, sites, checks, incidents and alert destinations, and never includes access keys or passwords. For anything else we hold about you, such as help questions or billing records, email support@getfeder.com. An Admin can delete the account on the same page with Delete my account (cancel a paid plan that is set to renew first). You can also request access to, correction of, or deletion of your personal data by emailing support@getfeder.com. We honor these requests in the spirit of the California Consumer Privacy Act (CCPA/CPRA) for our California customers: we will confirm what we hold, correct inaccuracies, and delete on request, subject to the legal record-keeping noted above.
We reply within 3 business days and never discriminate against you for making a request.
8. Security
We protect your data with measures including:
- Encryption of sensitive credentials (such as OAuth tokens and API keys) at rest.
- Encrypted connections (TLS) for data in transit.
- Strict separation between customer workspaces, so one agency cannot see another's data.
- Least-privilege internal access: only the systems and people that need access have it.
No system is perfectly secure, but we treat your clients' tracking data as the trust-sensitive asset it is.
9. Changes to this policy
If we change this policy materially, we will notify your account email at least 30 days before the change takes effect.
10. Contact
Privacy questions or requests: support@getfeder.com
Optional journey analytics
Journey analytics is an optional choice, separate from product usage counts: allowing it counts selected page visits and product activity for up to 7 days, while identifier-free page-open counts (route category and a coarse time bucket, plus viewport and referrer categories only when allowed) are kept for up to 30 days either way.
Also allow occupancy grid collection: coarse pointer or tap cells, scroll milestones, and foreground time. No text, form contents, or keystrokes. Occupancy counts are kept for under 7 days.
If you allow content experiments while a scan-results trial is offered, Feder can measure those trial signups. Experiments do not change prices, billing, or consent.