Who actually owns your district's family data? A contract-term scorecard
A district does not own its family data because a vendor says the district owns its family data. It owns it to the extent the contract says what happens on export, on termination, and on acquisition — and to the extent something stops the vendor selling to the families the district enrolled. This scorecard grades that language, from public documents only, against a rubric published before any vendor was scored.
| Dimension | Weight | Score 0 | Score 2 | Score 4 |
|---|---|---|---|---|
| Portability and export | 30% | No export term at all, or "reasonable assistance" with no format, deadline or fee cap | Export named as a right, format or deadline unspecified, fee possible | Named machine-readable format, named deadline, no fee, exercisable at any time and not only at termination |
| Deletion on termination | 25% | Silent, or deletion at vendor discretion, or indefinite retention of "aggregated" family data | Deletion promised on written request, no deadline or no coverage of backups and subprocessors | Deletion on a named clock, covering backups and subprocessors, with written certification to the district |
| Change of control | 25% | Assignable to any successor without notice or consent | Notice given, no exit right, terms may change on assignment | District may terminate without penalty on acquisition or material data-terms change, with export and a prorated refund |
| Family-facing monetisation | 20% | Vendor may market, advertise or sell subscriptions to district families; no restriction in the district paper | Restricted by privacy policy or pledge only, changeable unilaterally, or limited to "no sale of data" | Contractually barred from marketing to, advertising to or charging district families, for the life of the agreement |
Scores of 1 and 3 are interpolations between the anchors and require a written justification in the clause library. Weights were set before scoring began and will not move between editions; if they ever do, both weightings are published side by side.
Scored from documents, so nobody can decline to participate
Vendor questionnaires produce the scores vendors want. This is scored entirely from material the vendor has already published or filed, which means non-participation is not a strategy.
- Vendor master services agreements and terms-of-service pages, as published, with a retrieval date on every quotation.
- Signed National Data Privacy Agreements filed in the Student Data Privacy Consortium's public registry, where thousands of vendor–district agreements and their state-specific exhibits are searchable by anybody.
- State and consortium DPA registries maintained by state education agencies and regional service agencies, which frequently hold the executed version rather than the template.
- App-store terms and family-facing terms of use, which are often materially different documents from the district paper — and where they differ, the difference is itself a finding.
- Published subprocessor lists, because a deletion promise that does not reach the subprocessors is not a deletion promise.
No NDA'd material is used. No vendor is asked for anything. Every named vendor still receives its own rows ten business days before publication and any response runs verbatim beside the score.
The export reality check
Export is the dimension where marketing language and contract language diverge most sharply, so the scorecard carries a separate table that ignores the marketing entirely and asks four questions of the contract:
- What format is promised? A named CSV or JSON schema, an API, or the phrase "reasonable assistance", which is not a format.
- On what clock? A stated number of days, or "promptly", which in a dispute means whatever the vendor's counsel says it means.
- At what cost? Free, capped, or at the vendor's then-current professional services rate.
- When can you use it? At any time during the term, or only in a termination window — which is to say, only when you have already lost your leverage.
Our own score, stated plainly, including where we lose points. Kastr's clause 7.1 gives a district the contractual right to its data in a machine-readable format at any point in the term, at no charge. That is a contract right, and it is real. What does not yet exist is self-serve export tooling — there is no button in the product that produces the file today, and until there is, exercising 7.1 means asking us and us doing it. We will score ourselves on the same rubric as everybody else and we expect to lose points on that gap rather than argue our way around it. Clause 9.4 bars us from marketing to, advertising to or charging your families for anything, ever, and clause 11.2 lets you leave within 90 days with export and a prorated refund if we are acquired or materially change our data terms.
The clause library, and why every score is quotable
A score with no clause behind it is an opinion. Every cell in the scorecard links to the clause library entry that produced it, and each entry carries the verbatim text, the document it came from, the URL and the retrieval date.
The design goal is that a district's counsel can verify any single cell in under a minute without contacting us, and can verify the whole row for their own incumbent in about ten. That is also the honest test of whether this is research or advertising: if a lawyer can check it faster than we can defend it, it is research.
Clauses are archived, never overwritten. When a vendor changes its terms between editions, the diff view shows the old language and the new one side by side with both retrieval dates. Terms changing quietly is a well-documented pattern in this category, and an archive is the only thing that makes it visible.
The four-clause rider, free to copy
The scorecard would be a complaint if it stopped at scoring. It ships with a plain-language contract rider, CC BY, that a district can attach to any vendor's paper to close the four gaps the rubric measures. In summary:
- Export. District data returned in a named machine-readable format within a stated number of days of a written request, at any time in the term, at no charge, with the schema documented.
- Deletion. Deletion of district data within a stated period after termination, extending to backups and every subprocessor, with written certification to a named district officer.
- Change of control. Written notice within a stated period of a change of control, and a district right to terminate without penalty inside a stated window, with export and a prorated refund.
- No family monetisation. No marketing to, advertising to, or charging of district families or students, and no family-facing subscription revenue derived from the district's roster, for the life of the agreement and after it.
It is deliberately vendor-neutral. Nothing in it advantages Kastr, and a district that attaches it to a competitor's paper and gets it signed has got the thing this research is actually for.
Questions people actually ask
Can a school communications vendor sell or market to my district's families?
Unless your contract says it cannot, quite possibly yes. Privacy policies usually restrict the sale of personal data; that is a narrower promise than a restriction on marketing, advertising or selling a paid parent upgrade to the families your roster identified. Those are different questions and the scorecard grades them separately, because a vendor can be entirely compliant with student data privacy law and still be charging your parents nine dollars a month.
What happens to our data if our communications vendor is acquired?
It depends on one clause, and most districts have never read it. The default position in a lot of vendor paper is that the agreement is freely assignable to a successor, with no notice requirement and no district exit right, which means the terms you diligenced can change hands to a company you never evaluated. A strong clause gives notice, a termination right without penalty inside a defined window, export, and a prorated refund.
What does a real data-portability clause look like versus a fake one?
A real one names the format, names the deadline, states that there is no fee, and lets you exercise it at any time rather than only at termination. A fake one says the vendor will provide reasonable assistance in transitioning district data, which commits to nothing measurable and is typically billed at professional services rates at the exact moment your leverage has gone.
How did you score vendors without their cooperation?
Everything is graded from documents the vendors themselves published or filed: terms pages, master agreements, signed National Data Privacy Agreements in public registries, app-store terms and subprocessor lists. No questionnaires are sent, so no vendor can improve its score by responding well or damage it by staying quiet. Named vendors do receive their rows before publication and any reply is printed verbatim beside the score.
What is the SDPC National Data Privacy Agreement, and does signing it settle these questions?
The NDPA is a standardised student data privacy agreement developed by the Student Data Privacy Consortium and adopted widely across US districts, usually with a state-specific exhibit. Signing it settles a great deal about privacy and security obligations. It does not by itself resolve export format and timing, change-of-control exit rights, or whether the vendor may sell a paid upgrade to your families, which is precisely why those four dimensions are scored separately.
Can we attach your contract rider to a vendor's own paper?
Yes. It is published under CC BY with no gate and no registration, it names no vendor, and it gives Kastr no advantage. Districts routinely attach riders to vendor paper during negotiation, and a rider that closes four specific gaps is considerably easier to get signed than a redline of the whole agreement.
One price. Every feature. Locked for three years.
$3.50 per student per year under 5,000 students. No tiers, no add-on modules, no per-message fees. Published on the site because you should not have to book a call to learn a price.