Privacy policy

First Instinct Publishing · last updated 5 September 2026

First Instinct Publishing is a private automation tool. It is run by one person, on their own computer, to publish and measure videos on a single YouTube channel that they own. It has no users other than its operator, no accounts, and no server.

This policy explains what Google and YouTube data the tool accesses, how it uses that data, where it is stored, who it is shared with, and how access can be withdrawn. It references and is governed alongside the Google Privacy Policy. By using an application built on YouTube API Services you are also agreeing to be bound by the YouTube Terms of Service; see the terms of use.

Whose data this is about

The only Google account this tool ever connects to is the operator's own — the account that owns the First Instinct YouTube channel. It does not connect to, read from, or write to any other person's account or channel. It collects no data about viewers, subscribers or commenters as individuals.

What it accesses

The tool uses YouTube API Services and requests exactly three permissions. Nothing wider is requested, and no permission is requested that is not listed here.

Scope	What it reads or does
`youtube.readonly`	The connected channel's ID, title, handle, subscriber count and video count.
`youtube.upload`	Uploads a video file, its title, description and thumbnail to that channel, with the visibility the operator selected — private, unlisted or public — and the Made for Kids designation the operator made. Private is the default when the approval is created; public requires a further explicit acknowledgement.
`yt-analytics.readonly`	That channel's own aggregate performance figures: views, watch time, average view duration and percentage, likes, comments, shares, subscribers gained and lost, traffic sources, and audience retention where YouTube provides it.

Analytics figures are aggregate numbers about the channel's own videos. They do not identify individual viewers, and the tool does not request, receive or attempt to derive any information about who watched anything.

How it uses that data

Channel identity is compared against the channel the tool is configured for. If they do not match, the run stops and nothing is uploaded. This is the tool's main safety check.
Upload access is used only after a person has reviewed a finished video, approved it by hand, selected its visibility, and designated whether it is Made for Kids. The program selects neither value itself: an approval missing either one is refused, not defaulted. There is no scheduled or unattended upload path. Immediately before an upload begins the tool displays the YouTube certification notice and the Made for Kids designation notice, and stops if certification is not given.
Analytics are recorded so the operator can see how an episode performed and decide what to make next. Figures YouTube does not provide are stored as unknown; they are never filled in with zeros or estimates.

The data is not used for advertising, not used to build profiles of anyone, not sold, and not used to train any model.

Where it is stored

Everything is stored locally, on the operator's own computer. There is no server and no cloud database. Credentials are transmitted only to Google's OAuth and API endpoints, as any client must in order to authenticate and to make authorised requests on the operator's behalf. They are sent to no other destination.

The OAuth refresh token is written to a single local file with owner-only file permissions. It is excluded from version control, and is never printed to the screen, written to a log, or copied elsewhere. It is sent to Google's token endpoint when a short-lived access token is needed, which is what a refresh token is for.
Access tokens are held in memory for the life of a single run and are never written to disk.
The channel ID is held in a local environment file, also excluded from version control.
Analytics figures are written to a local database file on the same computer.

No credential is ever committed to a code repository. An automated check runs before every upload and refuses to proceed if one has been.

Who it is shared with

Google and YouTube user data obtained through this tool is not sold, and is not shared with any unrelated third party. It is not transferred to any other application. The only destinations it is ever sent to are Google's own OAuth and API endpoints, which is where it came from.

To be precise rather than sweeping: the tool integrates no analytics provider, no error-reporting or telemetry service, and no advertising partner. It does run on an ordinary personal computer, using that computer's operating system, storage and network connection in the normal way, and those are supplied by their own vendors under their own terms. That is the honest boundary of what a program running on a laptop can claim.

First Instinct Publishing's use and transfer of information received from Google APIs to any other app adheres to the Google API Services User Data Policy, including the Limited Use requirements.

How long it is kept

YouTube's Developer Policies allow data retrieved through the YouTube Analytics API to be kept for as long as it is needed, on one condition: the client must confirm at least every 30 days that the user still authorises it to hold that data. The tool implements this.

Every successful channel verification records the date of confirmation. Normal use therefore keeps the obligation current.
There is no grace period. Once a confirmation is more than 30 days old and Authorized Data is still stored, the tool stops treating retention as authorised: it refuses to upload until either a fresh confirmation succeeds or the stored data is purged.
If no data is stored, an expired confirmation blocks nothing — there is nothing being retained.
If a token refresh fails because authorisation is gone, that is recorded as a revocation and the same requirement applies.

Deletion removes all data obtained from the YouTube APIs — performance metrics, retention points, traffic sources, ingestion records, the publication index and the stored channel baseline — and writes a dated deletion record. Episode plans and question data, which are the operator's own work and contain nothing from any Google API, are not affected.

Withdrawing access, and what each step actually does

These two actions are not the same, and it is worth being exact about the difference.

Deleting the local credential file stops this tool from obtaining any further access tokens. It does not revoke the grant at Google, and it does not invalidate an access token that has already been issued — such a token remains valid at Google until it expires on its own, typically within about an hour.
Revoking access at myaccount.google.com/permissions withdraws the grant at Google itself. This is the authoritative step, and it is the one to use if the intention is to end access rather than merely to stop using the tool.

First Instinct Publishing provides a command that does all of this. It revokes the grant at Google through Google's own revocation endpoint, deletes the local credential, and purges every row of stored YouTube data, in that order. It asks for a typed confirmation first, writes a dated audit record containing no token value, and reports accurately if revocation succeeded at Google but local deletion did not. If Google reports the credential was already invalid, that is treated as the end state rather than as an error. Episode plans, the question bank and rendered video are not touched — none of that came from a Google API.

The command is available whether or not the operator also uses Google's permissions page. Using both is belt and braces; using Google's page alone is always sufficient to end access.

Children

The tool is not directed at children and collects no information from them. The YouTube channel it publishes to is set at channel level as not made for kids.

Changes to this policy

If what the tool accesses, stores or shares changes, this page is updated before the change takes effect, and the date at the top is revised.

Contact

Questions or complaints about these privacy practices can be directed to the operator using the user support email shown on the Google consent screen for First Instinct Publishing. That screen is where the operator authorises this tool, and Google displays the support address there. Because the tool has exactly one user — its operator — that is the contact route, and there is no need to publish a separate address here.

There is no contact form on this site, no help desk, and no support team. First Instinct Publishing is a program one person runs on their own computer; a message about it reaches that person directly or not at all.