The questions schools and studios actually ask — answered straight.

The book is print-on-demand with a soft markup: your retail price is your real print cost times a markup you set, at or above the cost floor. There is no contract, no minimum order, and no overprint — you pay for books people actually buy. Picture-day portrait packages and recognition ads are tiered by what’s included and by size. The specific prices are owner-configured in the platform and shared during onboarding — and your exact book price comes from a live print quote against your trim and page count. See how pricing works or get a quote.

No per-seat license. We are monetized by a share of the transactions that already flow — a slice of book and portrait net, and the card-fee-only margin on ads — deliberately below what a single bundler contract extracts. For a studio that owns picture day, the software is free: the studio keeps the largest portrait leg and we earn our residual on the same orders. The exact take-rate is owner-set and shared during onboarding. We win on breadth and volume, not a seat count.

No. A cost floor is enforced in code: a retail price below the live print cost is rejected at checkout, so a book can never be sold at a loss — even if someone fat-fingers the markup. Promo codes discount on top but are capped at an owner-set limit, and the floor still holds underneath them.

That multi-year exclusive — the “Term” the Big-4 reps sell — is genuinely real, and it is their moat, not your benefit. Homeroom is no-contract, no-minimum, no-overprint by design, so you can run one school or one publication alongside your incumbent for a season and compare the numbers before you move anything. Print-on-demand means there is no leftover-inventory risk to running a small pilot. When your Term ends, the switch is a decision, not an escape.

Roster import is roster-and-SSO in one sitting: SAML, Google / Microsoft OIDC, or an LTI launch, plus PSPA / SIS imports for picture day. Pages are built in our editor rather than migrated pixel-for-pixel from a competitor’s proprietary format — which is honest: a true round-trip of another vendor’s book file is not something we (or they) offer. Most pilots start a fresh publication, or rebuild a section, precisely because there is no contract forcing an all-or-nothing cutover. Roster & SSO shipped

There is no lock-in on the way out either. Your finished publication exports as a print-ready PDF (with fonts embedded), and your roster and content are yours to take. We never reintroduce a forced-overprint or minimum-order commitment that would strand inventory on your shelves — that pain is exactly the wedge we were built against.

It stays inside the school’s walls. The picture-day pipeline runs in the school’s isolated cloud tenancy: no face is uploaded to a third-party AI service, and no biometric template is ever created — so there is no faceprint to leak, subpoena, or sell, because it does not exist. This is the opposite of the bundler model, where every child’s portrait egresses to a vendor cloud for facial-recognition matching. No-egress pipeline shipped

By the school’s roster, not a face match. A guardian proves their student through a claim code or a roster match — the same authoritative source the gradebook uses — and that resolves which portraits are theirs. It is a database query the school already trusts, not a surveillance feature. (An optional face-assisted convenience exists only behind the same consent gate, and is unavailable until we procure in-VPC models; the roster path works without it.) Roster lookup shipped

No. A studio, a rep, or a sales account reads zero student rows. That is not a policy promise — it is a restrictive database rule (the rep PII wall) proven against real Postgres across the full student-data census: a studio or rep session returns no student records even with the school in scope. Reps see aggregates, adult contacts, and reconciliation; never a child’s record. The same kind of rule enforces a single-school FERPA wall, so an adviser at one school cannot read another school’s students either. FERPA & rep walls shipped

Selling a likeness is a commercial use, which the school’s educational COPPA attestation does not cover — so a portrait of an under-13 or unknown-DOB student is unsellable until verifiable parental consent, captured naturally when the guardian claims their student and checks out. Every gate is fail-closed: no consent on file means no sale, enforced in code (the consent gate), not by a checkbox someone could forget. A do-not-publish or opt-out student is suppressed end-to-end. The binding terms and the data-controller relationship live in the in-app policy your administrator agrees to; the final compliance artifacts are in counsel review.

Honestly: a Data Processing Addendum, a privacy policy, a COPPA disclosure, a security overview, and a sub-processor list all exist as engineering-grounded drafts in counsel review — they are not yet final, and we will not pretend otherwise. Our accessibility conformance is currently code-review-grounded, with an independent audit on the path to an externally-attested VPAT. We’ll share the current drafts and the review status up front, because your counsel signing our DPA is a real step we plan for, not paper over.

A studio that runs picture day keeps the largest leg. A portrait order settles on net and splits four ways — studio, photographer, school, and platform — with each share pinned at payout creation on an append-only ledger; the studio keeps the biggest share by design and the school keeps a fundraising leg. If you run picture day in-house instead of with an outside studio, that studio leg collapses into the school, so the school keeps the lion’s share — you are never the leg that gets squeezed. The exact percentages are owner-configured and shared during onboarding. This is the inverse of the incumbent model, where the vendor captures the whole portrait pool.

Lifetouch genuinely does run an end-to-end operation — that part is fair to them. The trade-off is that it “handles everything” by taking your data out of the building: portraits egress to a biometric cloud, and the parent-portrait sales profit is the vendor’s, not the school’s. With Homeroom the picture day stays inside the school’s walls, the school (or your studio) keeps the portrait profit, and your district’s counsel can sign a no-egress, no-biometric posture. That is the whole pitch — not that they do nothing, but that you keep the data and the money.

TreeRing genuinely ships per-student personal pages, a no-overprint/print-on-demand model, and a real fundraising story — credit where it’s due, and we match those: per-student personal pages, print-on-demand with no overprint, and fundraising that returns 100% to the school are all shipped here. What we add on top is the photography-revenue machine — the no-egress picture-day pipeline plus the parent-portrait split — and a multi-publication newsroom and teach-as-you-build curriculum that TreeRing does not own end-to-end. Personal pages & fundraising shipped

Live today: the desktop-class book builder (undo/redo, version history, layers, masking, auto-layout, brand-locking, full keyboard parity), the multi-publication editor (yearbook, newspaper, lit-mag, public online editions), the newsroom workflow (assignment → versioned submission → rubric review → editor-terminal approval), the ad-sales engine, fundraising (100% to the school), the no-egress picture-day pipeline and the consent gate, the FERPA and rep PII walls, and print-on-demand with fail-closed preflight and guaranteed font embedding. Shipped

In early access, building now on a foundation that is already live: the parent-facing portrait storefront and the four-way split (the no-egress pipeline and the consent gate underneath it are shipped — the storefront is the front door being built on them), buy-now-pay-later installments at checkout, and a cooperative-purchasing vehicle for districts that want to skip an RFP. We show these as the near-term plan with the hard half already built, never as live. Early access

Open it and watch — we’d rather show than claim. Undo/redo, version history, layers, masking, auto-layout, brand-locking, and full keyboard parity are shipped and tested. And it is one tool, not the novice-tool-vs-pro-tool split the incumbents ship (a simple online designer for most users and a separate Adobe-grade path for the few). We’re also honest about the two real remaining gaps: in-frame photo crop and a tunable drop-shadow are building now. Editor shipped Crop & shadow building

We’re honest about this rather than impressive about it. Real generative AI (captions, alt-text, theme suggestions) is on only with an API key, and it fails loudly when the key is absent — it never fakes canned text and passes it off as real AI. Face-detect and one-click background cutout are deterministic stubs until we procure the GPU and wire the in-VPC models. We’d rather tell you exactly where each piece is than ship a convincing fake.

For a pilot we keep the loop small and observable: one studio (or one school) and one to three publications, with a single point of contact who does the platform setup while your adviser owns the roster, the consent capture, and the book. Provisioning a studio or a school is a shipped self-service flow, and a school can be onboarded — roster, SSO, brand — in one sitting. We define your success metrics with you before go-live so renewal is a number, not a vibe.

No multi-year exclusive, no auto-renewing “Term,” no minimum-order commitment, and no forced overprint. A pilot is a short, plain agreement for one to three publications. The compliance paperwork your district needs — the DPA in particular — is in counsel review, and your counsel signing it is a planned step we’ll walk through, not something we’ll rush you past. You stay because the economics and the tool are good, not because a contract traps you.

Honestly, not yet — but it’s the near-term district unlock, not a live feature. A cooperative-purchasing vehicle (a lead contract a district can piggyback) plus enrollment-tiered pricing is on the roadmap so a district can skip its own RFP. We name it as building, not shipped. Cooperative purchasing in early access