What we ask for, and what we never ask for.

What we collect

Only what the research needs: your target area, the school year and grade you are planning for, the school types you want compared, your budget and funding situation, commute and aftercare constraints, sibling schedules as logistics, and what you tell us about your child in your own words.

We also collect the email address and billing name attached to your payment, which our payment processor handles.

What we never ask for

• Medical or diagnostic records. If a learning or support need matters to the research, describe it in your own words. We do not need documentation.
• Social Security numbers. Not for you and not for your child.
• Full school records. No transcripts, discipline files, or cumulative folders.

If a form ever seems to ask for one of these, stop and write to us. It is a mistake on our side.

Your child's information

We deliberately keep this thin. The report covers one focus child, and the only child information we hold is what you type into the intake and any clarification you send us. We do not ask for a photograph, a birth date, a student identification number, or contact details for your child.

The AI report guide on the homepage

The optional homepage guide explains the report and its limits. You can explore the example questions without AI processing. Typing sends your question to OpenAI; choosing Talk streams your microphone audio through LiveKit to OpenAI. Please do not share child names, diagnoses, records or exact addresses. You may optionally share an adult preferred name during voice; it is used for the current conversation and is not saved as a profile. The guide cannot access your private intake or order.

For this guide, we request written answers without saving them in OpenAI's application response history. We do not save recordings or transcripts for this guide. OpenAI may retain content in abuse-monitoring logs, normally for up to 30 days and longer for legal or safety reasons. Deleting your information from School, Clearly does not necessarily remove those provider logs. The private intake helper below has its own processing and storage description.

The visible conversation stays in the page's temporary memory. Minimizing the guide, pressing Escape, or opening the sample or fit check pauses voice and keeps the conversation on this page. Clear conversation or reloading the page clears it. Follow-up answer selection uses the last six public answer references and the selected example step. During voice, the latest guide reply also provides temporary context for your next question. That reply is held in worker memory, sent to OpenAI with the next question, and cleared when voice stops. No visitor transcript history is added to these answer requests. Nothing transfers into your private intake. We mask the guide's content from session replay and do not send questions or answers to website analytics.

We keep content-free session and usage records to limit abuse and account for costs. A protected hash of the network address limits repeat starts. The address hash and session credential are cleared within 25 hours after the session expires. Anonymous usage and cost records contain no question, answer or audio. You can mute the microphone, mute the speaker, or stop voice at any time. Muting the speaker leaves the microphone and connection active; Stop voice stops both.

The optional live voice helper

You can always type your answers. If you prefer, an optional live voice helper lets you speak them in a live back-and-forth instead. This is not the same as recording a clip, described in the retention section below. When you turn it on and agree, it streams your microphone audio to our AI provider as you speak, so it can suggest answers back to you. You can mute it or stop it at any time, and nothing you say enters your saved answers until you review the suggestions and choose Apply.

Two providers make this work. OpenAI runs the helper: its GPT-Live-1 model handles the live voice, and its GPT-5.6 model drafts the suggestions you review. LiveKit carries the real-time connection. We do not store your voice recording or a raw transcript of it. OpenAI does not keep the live session as a saved recording. Its backend response logs retain the text used to prepare suggestions and the resulting responses for at least 30 days, subject to its retention exceptions. Separately, it may retain conversation content in abuse-monitoring logs for up to 30 days, or longer when legally required or needed to protect people or services from harm. These provider logs are separate from the copies School, Clearly controls, so deleting your information from our service does not necessarily remove those logs. LiveKit is configured to keep no audio, transcript, or trace, and the session room is deleted when the call ends. We temporarily store encrypted suggestions so you can review and apply them. Those temporary suggestions are deleted when the session finishes closing. Answers you apply become part of your saved intake and follow the schedule described below.

How long we keep it

We delete your intake, our research working files, and our copies of your bundle 30 days after we complete delivery or resolve your last correction, whichever is later. That covers the 14-day correction window in our terms with room to spare. Download your files when they arrive, because after 30 days we can no longer resend them.

Recorded clips go sooner. If you record a short audio clip for one answer and upload it, we transcribe it and delete the raw audio within seven days of your confirming what it said, or within 30 days of upload if you never confirm, whichever comes first. We keep only the confirmed text.

The live voice helper described above works differently. It streams your audio as you speak and leaves us no recording and no transcript, so there is nothing of that kind to delete. Temporary encrypted suggestions are deleted when the session finishes closing. Answers you review and apply follow the same schedule as the rest of your intake, above.

What survives is a plain order record: your order reference, your email address, what was delivered and when, and the bundle version. It holds nothing about what your child is like. We keep it as long as tax, accounting, and payment-dispute rules require. Card details sit with our payment processor on their own schedule and never with us.

You can ask us to delete your information sooner, and you can ask for a copy of what we hold. Write to the address below and we will confirm when it is done.

Who sees it

Eric reviews the research behind every report. Beyond that, your information is handled only by the services we use to run the business, which at present are our payment processor, our intake form, our email provider, the AI tools used to draft and check research, and, when you use the optional voice helper, OpenAI and LiveKit as described above. Those services process your information on our instructions.

We do not sell your information. We do not build a profile of your family for any purpose beyond producing your report. If you buy a report, we tell Meta that a purchase happened, as described below.

How we measure the website

This site uses Google Analytics to count visits and see which pages families read. It tells us how many people arrived, which page they landed on, roughly which city they came from, and what they clicked. It does not tell us who you are, and it is separate from anything you type into an intake form or send us by email.

Google sets a cookie in your browser to do this. We have turned off the settings that would share this measurement for advertising or use it to build an advertising profile, and Google Analytics is not connected to our advertising. We also switched off form interaction tracking, so nothing you type into the calculator reaches Google. We keep this website measurement for 14 months. Any browser setting or extension that blocks Google Analytics will stop this site from measuring your visit, and every page works the same either way.

We also use Microsoft Clarity to see how a page is actually used: how far down people scroll, what they click, and where a layout confuses them. Clarity rebuilds the page from its own structure rather than filming your screen, and it hides sensitive content before it leaves your browser. Everything typed into a box and every dropdown selection is hidden that way. We went further on the tuition calculator and marked the whole tool, the entry fields and the results it works out, as content Clarity must not capture, so neither the figures you enter nor the figures we show you leave your browser. It does record which buttons and options you select, including the three fit check answers on the home page.

Clarity sets a cookie so the pages you view in one visit stay linked together. Microsoft keeps a recording for 30 days, and up to nine months for a randomly chosen sample and for any recording we save to come back to. It is not connected to any advertising account, and as with Google Analytics, blocking it changes nothing about how the pages work.

We advertise on Facebook and Instagram, and we tell Meta when someone buys a report so we can see which ads work. This happens on our server after your payment completes, not in your browser, and there is no Meta pixel or tracking script anywhere on this site. We send three things: your billing email address scrambled into an unreadable code, which Meta compares against its own scrambled codes and cannot turn back into your address; the amount of the purchase; and the fact that it was a purchase on this website. We do not send your name, your child, anything you typed into an intake form, the fit check or the calculator, or anything about the pages you read. Nothing is sent if you do not buy. To object, write to hello@schoolclearly.com and we will stop sending your purchase, and Meta's own privacy policy describes the settings it offers you directly.

If you sign up for school updates

You can read everything on this site without subscribing. For Houston school updates, the standard page form asks for one thing: your email address. In the homepage guide's signup card, your first name is optional and is used to greet you in emails. Both forms ask for your own adult email address, with no child field or questions about your family.

Our email provider MailerLite runs the standard page form. We use single opt-in: choosing the signup button subscribes you, and new subscribers receive a welcome email without a separate confirmation click. The standard form also carries a Google reCAPTCHA check, which is there to keep automated signups out, and Google processes the technical information that check needs, such as your network address and browser.

In the guide, you enter your details, review and edit them, then choose Sign me up. Voice stops before the contact card opens. Those fields are never sent to the AI providers, included in conversation history, or saved in persistent browser storage. Closing the card, minimizing the guide, clearing the conversation, or leaving the page clears those fields. A request already submitted can still finish. Any adult name shared during voice stays separate and does not fill the signup form.

The chat signup uses Cloudflare Turnstile to check for automated requests. Cloudflare processes the technical browser and network information needed for that check. Our server sends the reviewed email and, for a genuinely new record only, the optional name to MailerLite. A later signup never replaces an existing subscriber's name. To limit duplicate or abusive requests, we keep keyed codes derived from the address, network address and reviewed request, with its random request ID, time and outcome. These receipts contain no readable email, name or security token and are removed within 48 hours. Contact fields are masked from session replay and excluded from website analytics.

MailerLite holds your email address, any optional first name saved for your subscription, your signup record and any earlier confirmation record, whether you are currently subscribed, and which links you click. We have email-open tracking switched off. Replies come to our business mailbox and a person reads them, so please keep child names, records, and private family details out of them. MailerLite's privacy policy describes its own processing.

Every newsletter carries an unsubscribe link, and it stops the email immediately. To have the subscription record itself removed, write to hello@schoolclearly.com. We delete it from MailerLite and tell you when it is done. We tested that deletion on September 12, 2026 and the record was gone afterwards. Removing a subscription does not reach MailerLite's own service records, which follow its policy linked above. We keep your subscription information for as long as you stay subscribed.

This is separate from everything else on the site. Nothing you type into the tuition calculator, the fit check, or an intake form is sent to MailerLite. Buying a report or writing to us for help does not subscribe you to anything. If you buy a report using an email address already in MailerLite, we mark that existing subscription as a report buyer so the welcome follow-up emails stop. This does not create a subscription or change your subscription status.

We do not contact your child's school

No step in producing your report involves contacting a school your family is currently at or applying to. The action kit includes an email you can send yourself, so that any approach to a school comes from you.

Who we are and what this is

School, Clearly is AI-assisted research and decision support with human review. It is not placement advice and not a credentialed educational recommendation. Your family owns the final school decision. Admission, placement, and fit are not guaranteed.

No school pays us. We take no referral fees and no commissions from any school named in your report.

Questions or corrections

Write to hello@schoolclearly.com. If something in your report looks wrong, tell us and we will check the source and correct it.