The State of District Communication Procurement, 2027
This is the one asset in the programme that is not built from documents, which means it carries the most methodological risk and needs the most disclosure. The instrument is published before fielding opens, the microdata is released with the report, and the sponsor of the survey is a vendor with a commercial interest in the answers. All three of those facts belong on this page rather than in a footnote.
| Module | Core question | Reported as | Core or rotating |
|---|---|---|---|
| Procurement route | RFP, cooperative contract, sole source or renewal without solicitation | Distribution, cut by enrolment band | Core |
| Price awareness | What do you believe you pay per student? | Belief versus the district's own public contract where matchable | Core |
| Re-selection | Would you select your current platform again? | Share saying no, cut by tenure and procurement route | Core |
| Switching friction | What has stopped you changing platform? | Ranked barriers plus verbatim open text | Core |
| Channel mix | Which channels carry which message classes | Distribution, cut by region | Core |
| Language access | How translated notices are produced and reviewed | Distribution plus open text | Rotating, edition one |
| Contract terms | Awareness of export, deletion and change-of-control terms | Share who have read each clause | Rotating, edition one |
| Respondent frame | Role, enrolment band, region, district-domain verification | Response distribution and non-response discussion | Core |
A fixed core plus a rotating module means edition one is already the first point of a trend line rather than a one-off. Core wording does not change between editions; if it ever must, both wordings publish side by side and the series breaks visibly.
Recruitment, verification and incentive
The failure mode of vendor-run surveys is a self-selected sample of the vendor's own users, reported as though it were the sector. We cannot make that mistake, for a slightly awkward reason: we have no users. There is no customer base to contaminate the frame with, and stating that plainly is more informative than any weighting scheme.
- Screening on a district domain. Respondents verify with a district or educational email address. Responses that fail verification are excluded and counted in the exclusion ledger.
- Recruitment through associations and trade lists — CoSN, CETPA, TCEA, state CTO groups, NSPRA chapters, and trade-press reader lists including K-12 Dive, District Administration and eSchool News. Every recruitment channel is named in the report with the share of completes it produced, because channel composition is most of what determines who answers.
- The incentive is a $10 DonorsChoose donation per completed response, not a gift card. It removes the personal-benefit question that a district ethics officer would otherwise have to think about.
- Non-response bias is discussed rather than assumed away. A survey about procurement dissatisfaction attracts the dissatisfied, and any honest report of a regret figure has to say so next to the figure.
Margin of error publishes with n and with the frame it applies to. Where a cut is too small to support a margin, the cut publishes with its n and no margin rather than with a spuriously precise one.
The price-awareness question, which is the interesting one
Respondents are asked what they believe their district pays per student. Where the district can be matched to a contract in the pricing index corpus, belief is compared to the document.
The gap between those two numbers, in either direction, is the finding this survey exists for. It is measurable, it has never been measured in this category, and it is the empirical version of the argument this whole site makes: that a market without published prices produces buyers who do not know what they are paying.
Matching is done at the district level and reported only in aggregate. No individual respondent is ever identified, and no district is named alongside a belief figure. Where matching would make a respondent identifiable — a single district in a thin cell — the row is suppressed.
Microdata release, and the pre-registered instrument
The anonymised response-level dataset and its codebook are released at the same moment as the report, not afterwards and not on request. Anyone who wants to re-cut the data by region, by enrolment band or by role can do so without asking us, including anyone who suspects the report of putting a thumb on the scale.
The full questionnaire goes up before fielding opens, on this page, with a version number. Question order, scale anchors and any skip logic are all visible in advance. It is not possible to write questions backwards from a conclusion if the questions were published before the data existed, and that is the only defence a vendor-sponsored survey has that is worth anything.
The conflict of interest, stated in full. Kastr is a communications vendor. We benefit commercially if this survey finds that districts are dissatisfied with incumbents, unaware of what they pay, and locked in by switching friction. That is a real conflict and no methodology eliminates it. What can be done about it: publish the instrument first, publish the microdata, name every recruitment channel, discuss non-response openly, and invite anyone to re-run the analysis and contradict us. If the survey finds districts are broadly satisfied and well informed, that result publishes with the same prominence and the same press outreach.
Questions people actually ask
How do districts actually choose a communications platform: RFP, cooperative contract or sole source?
All four routes are in use and the mix varies sharply by enrolment band, which is why the procurement-route module cuts by size. Large districts tend toward formal solicitation; smaller ones lean on cooperative contracts that satisfy a competitive-procurement requirement without a local RFP, and a substantial share of renewals happen with no solicitation at all. The survey measures that distribution rather than repeating the folklore about it.
What share of districts would not choose their current platform again?
That is the headline the survey is designed to produce and we will not guess at it in advance. When it publishes it will carry n, the margin of error, the cut by tenure and procurement route, and an explicit discussion of the fact that a survey about procurement attracts the dissatisfied.
Do district leaders know what they pay per student?
This is the question we most want answered. Respondents state what they believe they pay, and where their district appears in the pricing index corpus, belief is compared with the contract. The gap in either direction is the result, reported in aggregate only, with suppression wherever a cell would make a respondent identifiable.
Who was surveyed, how were they recruited, and what is the margin of error?
District technology directors, communications directors and business officials, verified against a district or educational email domain, recruited through association and trade-press channels that are each named in the report with the share of completes they produced. Margin of error publishes with n and the frame it applies to; cuts too small to support one publish with their n and no margin instead of a false precision.
Can I get the raw response data?
Yes, immediately and without asking. The anonymised microdata and codebook are released simultaneously with the report under CC BY. Releasing the rows is the part nobody in this category does, and it is the only way a reader can check whether the headline was cherry-picked.
Does Kastr's sponsorship of the survey bias the findings?
It is a real conflict and we would rather name it than manage it quietly. We benefit if districts turn out to be dissatisfied and uninformed. The mitigations are procedural: the instrument publishes before fielding, the microdata publishes with the report, recruitment channels are named, non-response is discussed, and any finding that is inconvenient for us publishes with the same prominence as one that is not.
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.