Plain-language privacy.

When you sign up and use the program, we store:

• Account: email address (your sign-in identifier).
• Payment: handled by Stripe — we never see your card number. We store the Stripe customer + session identifiers so we can match a purchase to your account.
• Program activity: which Steps and Parts you've completed, your Self-Assessment answers, your tasks (the to-dos you save), your Steady Work responses, and your Journal entries.
• Assistant use: when you use the assistant, Break it down, or the list builder, we keep how many times you used it and how much text the model processed, as numbers only, never what you said or what it answered. We keep these numbers for 400 days so we can see what the assistant costs to run.
• Calendar sync: only if you connect Google Calendar, we store the connected Google account email and encrypted OAuth tokens so we can mirror scheduled tasks to your calendar. To detect reschedules, we also read the minimum Google event metadata we need for sync — event IDs, status, start time, update time, and private extended-property metadata. We ignore unrelated events and do not store their details.
• Operational logs: error or mobile crash reports, which stay on infrastructure we operate (self-hosted GlitchTip). We use anonymized enrollment identifiers — never your email, never the body of a Journal entry, task, or Assessment answer. Mobile crash reports exclude request payloads, screenshots, and view hierarchy. Our public marketing pages use Umami and Plausible (both self-hosted, cookieless) to measure visits, traffic sources, and page interactions separately from your program data.
• Product analytics and session replay: while you are signed in to the program on the web or in the iPhone app, we use PostHog to see where people get stuck, hit an error, or get frustrated, so we can fix it. PostHog records which screens you visit, where you tap or click, and a replay of the screen that is masked: everything you type, and the text of your tasks, Journal entries, Steady Work, and Assessment answers, shows as blocked-out shapes, never words. Program lessons and the app's menus can appear as written. Recordings are tied to your anonymized enrollment identifier, never your name or email, and kept for up to three months. Signed-out pages, the Steady Check, and private request pages are never recorded.

The therapy page lets you request an initial assessment with Kevin through New Tree Counseling, his separate private practice. You do not need a Steady account. A request is not a confirmed appointment until Kevin approves it.

We store your name, email, optional phone number, requested time, telehealth or in-person choice, request status, and notification delivery status in a separate practice scheduling record. Kevin supplies a private meeting link when approving telehealth. Please do not enter symptoms, health history, or clinical records here. These records are not linked to your Program work.

Only practice-authorized dashboard accounts can read the contact details; being a Steady Admin alone does not grant that access. Your private link lets its holder view, change or cancel that request. Keep it private. Private request pages have no analytics or session recording, and booking data is excluded from crash reporting. Analytics on the therapy page count that a request form was submitted, never what you entered in it.

Our email provider, Resend, processes your email address and a private status link. Notification emails do not include your name, appointment time, format, phone number, meeting link or health history. Scheduling records remain on infrastructure we operate and in its backups. No information is sent to SimplePractice.

To request a correction or deletion, contact Kevin at kevin@newtreecounseling.com. Practice records are separate from Steady account deletion and remain subject to the practice's applicable retention obligations.

The Steady Check at /quiz is a self-reflection questionnaire you can take before you have an account. Here is exactly what happens with it:

• What we keep: your email address, optional first name, and your trait profile (per-trait averages and an overall figure). We do not store your individual question-by-question answers — they are scored and discarded in the same request.
• How long: quiz records are deleted automatically after 30 days.
• Email: the address you enter is where we send your results and the link to Kevin's free video. We may also occasionally send practical ADHD resources, program updates, and an invitation to join the Steady beta — the consent note on the quiz gate covers this, and you can unsubscribe at any time.
• Who touches it: the results email is delivered by Resend, our email provider. Everything else stays on infrastructure we operate. No advertising pixels, no session recording, and no third-party analytics run on the quiz pages.
• Deletion: email kevin@newtreecounseling.com from the address you used and we will delete your quiz records ahead of the 30-day schedule.

On the mobile app you can dictate a to-do or a piece of feedback instead of typing it. The speech-to-text runs on your phone's own recognizer, provided by iOS or Android — Steady's servers never receive your audio. For a to-do, we store only the text you choose to save, and that text is a task like any other: private to you and never shared with the Steady team.

When you capture something in the mobile app, Steady can sort it for you — split a sentence into separate to-dos and name each one. To do that, the sentence is sent to Claude, an AI model made by Anthropic. Under Anthropic's commercial terms it is not used to train their models, and Steady does not store the sentence anywhere along the way or send it to analytics or error monitoring. The model only splits and names to-dos — it never assesses you and gives no advice. Nothing is saved until you review what it found and press Add. Steady does keep a count of how often you use this and how much text the model processed — numbers only, never the words — for 400 days.

When our servers can't be reached, what you capture simply saves as a to-do, unsorted.

The mobile app has a feedback button. When you send feedback, we store the text you wrote or dictated, the screen you were on (with any task or lesson identifiers removed), and the app version — nothing else from the screen. Feedback is meant for the Steady team and is read by us to improve the product. We may write a short reply — it appears only in your own app, and the person replying never sees who sent the feedback. Both your feedback and any reply are included in your data export and deleted with your account.

If you ask Steady to break a task into smaller steps, the text of that task is sent to Claude, an AI model made by Anthropic. When you ask for more steps, or for one step to be made smaller, the steps you already have are sent with it. Under Anthropic's commercial terms this is not used to train their models, and Steady does not store it along the way or send it to analytics or error monitoring. The model is used only to split the task into steps, never to assess you. Nothing is saved until you review the suggestions and choose what to keep. Steady does keep a count of how often you use this and how much text the model processed — numbers only, never the words — for 400 days.

When our servers can't be reached, no steps are suggested and you can add your own.

The hard rule, the one that makes the rest of this policy meaningful: Nothing in Steady lets Kevin, or anyone else on the Steady team, read your Journal entries, your tasks, your Steady Work responses, or your Self-Assessment answers. Our systems enforce this in code, backed by automated tests; it is not a matter of trust.

What the Steady team can see is engagement metadata: which Steps and Parts you've completed and when, when you were last active on web or in the app, and counts (how many tasks, focus sessions, or journal entries exist — never what any of them say), and monthly totals of how much the assistant is used (counts of calls and of text size it processed, never what you said). In a month when only a few people used it, those totals can describe one person's use. This is how we notice someone is stuck and improve the program; the content of your work stays sealed.

Your data lives on our own servers, hosted in a United States data center operated by IONOS. Account details and program activity are stored in our database; files (handouts, lesson assets, data exports) live in our own S3-compatible object storage. Sign-in and sessions run on our self-hosted identity server. Encrypted backups are stored with Amazon Web Services (AWS) in the United States. Connections between your browser and Steady use HTTPS.

We use the following sub-processors to deliver the program. Each handles a narrow slice of data; none receives more than they need.

• IONOS — cloud infrastructure (a United States data center) hosting our self-operated database, file storage, sign-in, and the app itself.
• Amazon Web Services (AWS) — encrypted off-site backups only (database dumps and a mirror of our file storage), in the United States.
• Stripe — payment processing, billing portal, invoices.
• Resend — transactional email (account setup and invite links, export-ready notifications, lesson-ready notifications, and private practice-booking status links as described above), and occasional news emails from the Steady team to our mailing list, Enrollees and the waitlist. Resend keeps the address of everyone who has had one as a contact, along with whether they unsubscribed.
• GlitchTip (self-hosted) — web error monitoring and mobile native crash monitoring on our own infrastructure; crash data never leaves servers we operate. GlitchTip's own session replay, mobile screenshots, and mobile view hierarchy capture are disabled.
• PostHog — product analytics and masked session replay of the signed-in program on the web and iPhone, hosted in the United States (see “Product analytics and session replay”). Receives screen names, taps and clicks, and masked replays tied to your anonymized enrollment identifier; never the words you write.
• Anthropic — sorts what you capture in the mobile app into to-dos and breaks tasks into steps (see “Sorting what you capture” and “Breaking a task into steps”); receives only the sentence being sorted, or the task and its steps, and, under Anthropic's commercial terms, does not use them to train its models.
• Google Calendar — optional two-way calendar sync only if you connect your Google account; receives the scheduled task text, time, and Steady link needed to create and update calendar events. Steady also reads limited event metadata from Google to reconcile changes, and does not store unrelated events.

Our database, file storage, sign-in, error monitoring, and web hosting run on software we operate ourselves — these are self-hosted, not separate third-party services. Marketing-page analytics use Umami and Plausible, which we also self-host. Product analytics and masked session replay inside the program use PostHog, listed above. If you do not connect Google Calendar, we do not send your program data to Google for calendar sync.

You can, at any time:

• Export a copy of everything we hold about you from /portal/me. The export is a JSON file delivered via a signed link.
• Delete your account from /portal/me. Soft-delete is immediate; you have 14 days to change your mind by emailing kevin@newtreecounseling.com. After 14 days, your account and all linked data are permanently purged.
• Cancel billing from /portal/me/billing (Stripe Customer Portal).
• Ask us questions — email kevin@newtreecounseling.com.

California and EU/UK residents: the rights above (access, deletion, portability) satisfy the corresponding CCPA / GDPR obligations. There is no sale or sharing of personal information for cross-context behavioral advertising.

When you delete your account, Stripe retains its own transaction history per its tax + fraud-prevention obligations. We delete our link to that history; Stripe keeps the underlying records under its own privacy terms.

We use a session cookie from our self-hosted sign-in (Zitadel) to keep you signed in. We do not use third-party advertising cookies. Our public marketing landing page uses Umami (self-hosted analytics) to measure visits, traffic sources, and landing-page interactions; it does not set analytics cookies. Our home page and therapy page also use Plausible (self-hosted analytics) to measure visits, traffic sources, outbound link clicks, file downloads, and form submissions; it does not read what you type into a form or set cookies. When you arrive from a tagged link (for example an ad or a newsletter), your browser keeps that link's campaign name for up to 30 days, so a signup on another page of the site is credited to it; it holds no name, email, or identifier, and it is stored with your signup only if you sign up.

Emails from the Steady team (news and updates, sent from news.steadywithadhd.com) count opens and link clicks: Resend adds a small image and sends links through a redirect. We see only totals for each email, such as how many people opened it, never who opened or clicked. Every one has an unsubscribe link, and unsubscribing stops them. Account emails (sign-in links and codes, password resets, invites and reminders) are never tracked.

Steady is for adults. We do not knowingly collect data from anyone under 18. If you believe a minor has signed up, email kevin@newtreecounseling.com and we will close the account.

If we change how we use data in a way that affects you, we will update the “Last updated” date at the top of this page and email signed-in Enrollees ahead of the change taking effect.

New Tree Counseling, P.C. — Kevin Barr, LMFT.
kevin@newtreecounseling.com