Privacy Policy

Herding Chickens (“we”, “us”, “our”) is committed to protecting your privacy. This policy explains what personal data we collect, why we collect it, who processes it, and what rights you have.

It covers:

• Our marketing website at herdingchickens.com
• Our Telegram accountability bot (early access / prototype). You must accept this policy in Telegram before the bot will store your data or run for you

Herding Chickens is a productivity tool. It is not therapy, medical advice, or a medical device. Please do not use it as a substitute for professional care.

Who is responsible

Until a formal company entity is registered, the Herding Chickens founders act as joint controllers of personal data processed under this policy. Day-to-day privacy contact: contact@herdingchickens.com.

1. Marketing website

What data we collect

This marketing website collects only the data you actively submit:

• Newsletter email address — if you subscribe via the Resources or Contact page forms.
• Early-access registration — email address and optional name if you join the waitlist before the app launches.
• Contact messages — your name, email address, subject (e.g. general enquiry, support, or partnership), and message text when you use the Contact form — including creator-partner applications sent from the Partner with us page.

We use optional Google Analytics 4 on this website if you accept analytics cookies. Until you accept, we do not load the Google tag. We do not run advertising pixels. The banner at the bottom of the page asks for that choice; we store it in your browser (see our Cookie Policy). You can change it on the cookie policy page at any time.

Why we collect it and the legal basis

• Newsletter and early-access sign-ups — to send you the Herding Chickens newsletter, product updates, and (if you register interest) to notify you when early access opens. Legal basis: your explicit consent (UK GDPR Article 6(1)(a)), given when you tick the relevant box(es) and submit the form. The waitlist and newsletter each have their own required consent box. The other list is offered as an optional box, off by default.
• Contact messages — to read and respond to your enquiry, including partnership applications. Legal basis: our legitimate interest in handling correspondence (Article 6(1)(f)), balanced against your rights. You may object to this processing by emailing us.
• Website analytics (optional) — if you accept, Google Analytics records page views and similar usage (pages visited, referrer, approximate location/device). Legal basis: your consent (Article 6(1)(a)). You can refuse or withdraw on the Cookie Policy page.

Where we store website data

Newsletter and early-access sign-ups are stored in a secure database on our server (SQLite). If configured, we also sync those contacts to Loops (our email marketing provider) to send campaigns and automated welcome messages.

Contact form submissions are delivered to our team by email via SMTP (secure email transport). We do not store contact messages in the website database; they are held in our email inbox unless we save them for follow-up.

How long we keep website data

• Newsletter and early-access emails — for as long as you remain subscribed or registered. You may unsubscribe at any time using the link in our emails, or at herdingchickens.com/unsubscribe. We will delete your data within 30 days of unsubscribe or a deletion request.
• Contact messages — typically up to 12 months in our email inbox, then deleted unless we need to keep a thread longer to resolve an ongoing matter (e.g. a partnership discussion).
• Google Analytics— if you accepted, Google retains aggregated reports according to the property's retention setting (we use Google's standard Analytics retention). You can stop new collection by refusing analytics cookies.

2. Telegram bot

If you use our Telegram bot, we store the data from that use on our servers. Telegram also processes your messages under Telegram's Privacy Policy; this section is about what we store.

You must accept this policy before the bot will work

The first time you start the bot, Telegram will show a short notice with a link to this page and an Accept control (a tick / yes button). You must accept before we will run the bot for you, and you must confirm that you are 18 or over. If you do not accept, you cannot use the bot: nothing you send us is stored, and no part of the bot will respond to you.

Until you accept, we store nothing except a pseudonymised version of your chat ID and a record that we asked you.

We keep a record that you accepted, when, and which version of this policy and of the terms you saw. If we change either in a material way, we will ask you to accept again before continuing.

Accepting this policy also covers two AI features: task breakdown, and automatic area-guessing. When you use breakdown, the relevant task (and any related text we include in the prompt) goes to our AI provider. When a task you send doesn't match one of the bot's keyword areas, it may also send that task's wording to the same provider to guess which area it belongs in, rather than always asking you — you can correct that guess on the spot. We already store your task either way; we store the steps or the area the provider returns, not a separate copy of the prompt. See AI task breakdown and AI area-guessing below.

What we store

Once you have accepted, we store the following — and nothing else:

• A pseudonymised chat reference — we do not keep your Telegram chat ID alongside your data. Every record is filed under a one-way code derived from it, using a key we hold separately from the database, plus one encrypted copy of the ID so the bot can find your chat again to send a nudge after a restart. We do not store, and do not read, your display name, username or phone number.
• Your tasks— the wording of each task as you typed it, its area tag, whether it is open or done, when you created and completed it, which one you picked as today's main, and any step list you generated. Task wording is stored in plain text so the bot can show it back to you.
• Activity records — a log of what kind of thing you did and when, so nudges, streaks and our own reporting work. It records events like “a task was created”, “a check-in was answered ‘still on it’” or “the balance screen was opened”. See what the activity log contains below for the full list — it holds no text you have written.
• Preferences you set — timezone, nudge cadence, quiet hours, and that you have completed onboarding.
• AI breakdown — when you use Break into steps (or the equivalent) and the feature is enabled, we send the task text (and any related prompt text) to our AI provider, currently Anthropic, and we store the steps it returns. The prompt itself is not kept separately — it is the task you already sent us. That data is used for the same purposes as other bot data: running the bot, internal reporting, and improving the tool.
• AI area-guessing— when a task you send doesn't match a keyword area and the feature is enabled, we send that task text to the same AI provider to suggest an area, and store its suggestion as the task's area tag — nothing else new. See AI area-guessing below.
• Policy acceptance — that you accepted this policy, and the time of acceptance.

What we do not store

We do not keep a transcript of your conversation with the bot.There is no message store. Anything you send that is not captured as a task — menu taps, commands, your answers during setup, a message that just says “ok” — is acted on and discarded. All that remains is an activity record noting that a screen was opened or a button pressed, with none of your words in it.

We also do not store your Telegram display name, username or phone number, and the bot does not read them.

The bot cannot read photos, voice notes or files, and does not download them. If you send one it replies to say so. All that is recorded is that a message arrived and what kind it was — a photo, a voice note, a file — never the thing itself, and never a caption.

We do not ask for diagnoses or health information. Tasks might still mention sensitive topics — please only write what you are comfortable storing with us. We design the product as productivity software, not a health app. If you include that kind of detail, it is stored as part of the task you sent, for the purposes below.

What the activity log contains

The activity log is a separate record from your tasks: it is behavioural, timestamped, and filed under the pseudonym rather than your account. It never contains text you have written. In full, it records:

• each time you start the bot, flagged as first visit or return
• which screens you open — your list, today, a task's menu, its steps, balance, settings and each settings pane, the pickers for choosing or changing a main, for marking done and for breaking a task down, the rename prompt, and the timezone chooser
• each check-in answer — still on it, not yet, or done
• when you create or reword a task: how long it was and how many words, never the words themselves
• area tags applied, steps ticked off, mains chosen and resolved, and undos
• that you accepted or declined these terms, and which version you saw
• things the bot declined to do and why — a task over the length limit, a breakdown or an area-guess you asked for too often, an unrecognised chat, a message sent before you accepted, a breakdown our AI provider refused, or a message the bot cannot read because it is not text
• breakdown or area-guessing failures, recorded by error type only
• how far you got through setup, and where you stopped
• whether you kept or wiped your data when asked
• errors, recorded by error type only — never their message text

Each entry carries the time in UTC and your timezone offset, so we can tell what hour of your day it was without knowing where you are.

Why and legal basis

We use this data to:

• Run the bot for you— store your list, keep the conversation going, remind you about today's main, and keep settings and streaks accurate. Legal basis: this is necessary to provide the service you asked for after you accept this policy (UK GDPR Article 6(1)(b)).
• Internal reporting and improving the tool — understand how the product is used, find bugs, see what helps people start, and decide what to build next. That includes reviewing tasks and AI-generated step lists where they exist. Reporting is internal: we do not sell this data or use it for advertising. Where we can, we look at counts and patterns rather than naming an individual. Legal basis: our legitimate interest in operating, securing, and improving the product (Article 6(1)(f)). You can object by emailing us. If you do not want this processing at all, do not accept in Telegram — the bot will not run without that acceptance.

Where bot data is stored

Bot data is stored in a database on our application server / VPS. We aim to host in the UK or EU where practical. Access is limited to people who need it to run, support, report on, and improve the service.

AI task breakdown

The bot can split a task into smaller steps using an AI provider. When you use Break into steps (or the equivalent command) and that feature is enabled:

• We send the task text, and any related text included in the prompt, to our current AI provider, Anthropic, so it can generate the steps.
• Anthropic processes that content under its Privacy Policy. It may process data in the United States. We rely on appropriate safeguards for that transfer (such as Anthropic's data processing terms and standard contractual clauses) as required by UK GDPR.
• We store the steps returned in our database, together with your other bot data. The prompt is the task you already sent us (plus any related text included for the model) and is not kept as a separate record. We use the stored steps to show them back to you, and for internal reporting and improving the tool — the same purposes as your tasks.

If you never use breakdown, we do not send that task to Anthropic for this reason. Deleting your data with us (Start fresh or email) removes our copy; we cannot control any copy Anthropic may retain under its own retention rules.

By accepting this policy in Telegram you agree to that AI processing when you choose to use breakdown.

AI area-guessing

Every task you capture is filed under one of a small set of life areas (work, family & friends, health, personal). Most of the time this happens locally, by matching keywords in what you typed — no AI involved, nothing sent anywhere. When a task doesn't match a keyword and the feature is enabled:

• We send that task's wording to our current AI provider, Anthropic, so it can suggest an area. If it can't say confidently, we fall back to asking you directly, exactly as we always have.
• We store the area it suggests as the task's area tag — the same field a keyword match or your own tap would set. Nothing new is stored beyond that; we do not keep a separate copy of the prompt or the model's answer.
• The bot shows you the guess and a short window to correct it — until your next message or button tap, after which the guess stands (you can still recategorise indirectly by rewording the task in a way that matches a keyword).

By accepting this policy in Telegram you agree to that AI processing for tasks that don't match a keyword area, if the feature is enabled. If you'd rather it never happened, always phrase a task so a keyword area matches, or pick the area yourself when asked.

How long we keep bot data

• Twelve months after your last activity, we delete everything we hold for you— tasks, step lists, activity records, settings and your acceptance record. “Last activity” means the last time you sent the bot anything or tapped one of its buttons, so an account you keep using is kept; one you walk away from is cleared a year later.
• While you are using it, we keep your tasks, step lists, activity records, settings and acceptance record so the product works and so we can report on and improve it.
• You can wipe your bot data at any time with Start freshin Settings (two-tap confirmation). That permanently deletes your tasks, step lists, streak, settings, activity records and acceptance record. One marker is written afterwards recording that a reset happened, under the pseudonym only. Aggregated figures that no longer identify you (for example “how many people completed a main that week”) may remain.
• You can also email contact@herdingchickens.com to request deletion; we aim to complete erasure (including applicable backups where reasonably practicable) within 30 days.
• Deleting the Telegram chat alone does not delete our server copy — use Start fresh or email us.

Who we share bot data with

• Telegram — delivers messages between you and the bot.
• Hosting / VPS provider — stores the bot database and runs the service.
• Anthropic— when you use AI step breakdown, and when a task doesn't match a keyword area and area-guessing is enabled. We send the relevant task text so Anthropic can return steps or a suggested area. We store what it returns; we do not keep a separate copy of the prompt.

3. Who we share data with (summary)

Depending on which services you use, providers may include:

• Loops — newsletter and early-access list management and email delivery (when configured).
• Google LLC— Google Analytics 4, only if you accept analytics cookies. Usage data may be processed in the United States under Google's data processing terms.
• SMTP email hosting — delivery of contact-form messages to our team and optional transactional emails.
• Hosting provider — runs the website and/or bot and stores related databases.
• Telegram — messaging platform for the bot.
• Anthropic— AI breakdown of a task you choose to split, and area-guessing for a task that doesn't match a keyword: the relevant task text goes to Anthropic, and we store what it returns (steps, or a suggested area).

Loops, Google Analytics, and some AI providers may process data in the United States. Where personal data is transferred outside the UK, we rely on appropriate safeguards (such as the provider's data processing agreement and standard contractual clauses) as required by UK GDPR.

We do not sell your data to any third party and never will.

4. Your rights

Under UK GDPR you have the right to:

• Access the personal data we hold about you
• Correct inaccurate data
• Request deletion of your data
• Withdraw consent at any time (without affecting prior processing)
• Object to processing based on legitimate interests
• Lodge a complaint with the ICO (ico.org.uk)

To exercise any of these rights, email us at contact@herdingchickens.com. For bot data, you can also use Start fresh in the bot Settings. If you do not accept this policy in Telegram, we will not process your bot chat or tasks.

5. Children

The bot and site are for adults. You must be 18 or over to use the bot, and the bot asks you to confirm that when you accept this policy in Telegram. Telegram's own terms of service also require users in the United Kingdom, the EU and Australia to be at least 18 to hold an account. Do not use the service if you are under 18.

6. Contact

Herding Chickens · contact@herdingchickens.com