Choosing Between Text, Phone Call and Email
Every multi-channel article in this category discusses channel strategy at length and never once states a cost, because the companies writing them do not publish prices. Channel selection is an arithmetic problem before it is a strategy problem, and the arithmetic is not difficult.
| Scenario | SMS | Voice | Correct default | |
|---|---|---|---|---|
| Lockdown alert | Primary | Failover | Parallel | SMS, voice failover, email in parallel |
| Weather closure | Primary | Failover | Parallel | All three, staged |
| Two-hour delay | Primary | Yes | Detail | SMS + voice |
| Mid-day early release | Primary | Yes | Detail | SMS + voice, two waves |
| Bus running late | Primary | Too slow | Nobody reads it in time | SMS only |
| Same-day absence | Primary | Failover | Long form | SMS, voice failover |
| Attendance pattern conversation | Inadequate | Live call | Inadequate | A person on the phone |
| Truancy notice | Cannot serve | Heads-up only | Cannot serve | Paper, with an SMS heads-up |
| Health office pickup | Yes, no detail | Primary | No | Voice, SMS if unanswered |
| Communicable disease notice | Pointer | No | Primary | Email, SMS pointer |
| Immunisation deadline | Reminder | Final reminder | Detail | Email, then SMS, then a call |
| Meal balance | Careful wording | No | Primary | Email, SMS pointer |
| Registration window | Reminder | Once | Primary | Email + SMS reminders |
| Parent-teacher conferences | Reminder | Hard-to-reach only | Primary | Email + SMS reminder |
| Testing week | Night before | Yes | Detail | Email + voice + SMS reminder |
| Event reminder | Primary | No | Newsletter | SMS |
| Athletics change | Primary | No | No | SMS |
| Newsletter | No | No | Primary | |
| Survey | No | No | Primary | |
| Death of a student or staff member | Never first | Staff only | Families | Staff in person, families by letter |
The four that go on every channel simultaneously: an active safety incident, an all-clear, a same-day closure, and an unplanned early release. Everything else is a choice, and sending everything on every channel is how a district trains families to ignore all three.
The cost arithmetic, per 1,000 families
Use your own rates; the shape holds. Illustrative figures: SMS at 0.4¢ per segment, voice at roughly 1.2¢ per completed 60-second call, email at effectively zero marginal cost.
- SMS, English, one segment, 1,000 recipients: $4.
- SMS, accented Spanish, three segments, 1,000 recipients: $12.
- SMS to a 60/40 English-Spanish district, 1,000 recipients: $2.40 + $4.80 = $7.20 — nearly double the all-English figure.
- Voice, 60-second call, 1,000 recipients: around $12, plus the recording time of a human being.
- Email, 1,000 recipients: negligible.
- SMS with voice sent to everyone in parallel: $19.20 for the bilingual district. Roughly two and a half times SMS alone.
- SMS with voice as failover on the 6% that fail: $7.20 + $0.72 = $7.92. Ten per cent more than SMS alone, for materially better reach.
That last comparison is the single most useful line on this page. Failover costs a tenth of parallel sending and reaches nearly the same households, because the households that need the voice call are precisely the ones where the text failed. Kastr fails over from SMS to voice on terminal failure, on the same contact point, with the two delivery records linked so the pair is visible afterwards rather than looking like two unrelated attempts.
Scale it: a district sending 30 district-wide messages a year to 12,000 contact points spends roughly $2,600 a year on SMS-with-failover and roughly $6,900 on parallel SMS-plus-voice. The difference funds a staff position over a contract term.
What each channel is actually good at
SMS. Fast, near-universal, read within minutes, and terrible at length. Its weakness is that it carries no proof of anything beyond delivery to a handset and it cannot hold detail. Its hidden weakness is cost inflation on accented languages, which penalises exactly the families a district most needs to reach.
Voice. The only channel that reaches a landline household, the only one that works for a family with limited literacy, and the only one where tone communicates. It is slow to complete at scale, expensive per contact, and unusable for anything a family needs to refer back to. It is also the channel families most associate with something being wrong, which is a strength for a closure and a weakness for a book fair.
Email. Unlimited length, free, archivable, and the only channel where a family can read the whole thing twice. Open rates for district email are poor and getting worse, and it is useless for anything time-critical. It is the right home for the letter that follows the alert.
Language and accessibility by channel. A translated email or SMS is read in the family's language. A translated voice call is only as good as the recording, and a synthetic read of a language nobody on staff speaks should be reviewed by someone who does before it goes to twelve thousand households. Kastr's translation preview renders the draft through DeepL before you send so you can read it; for voice, get a human to listen to it.
Sequencing patterns, and when each is right
- Text then call, staged. Text goes immediately; the voice leg runs to the whole audience five minutes later. Correct when you need certainty and cost is secondary — a closure reversal, an all-clear after a serious incident. Expensive.
- Text with voice failover. Voice goes only to the contact points where SMS terminally failed. Correct for almost everything else, and roughly a tenth of the cost of the pattern above. This is Kastr's built-in behaviour.
- Call then text. Correct in exactly one situation: an audience skewed heavily to landlines and older contact records, such as a message to grandparent-guardian households. Rare, and worth knowing you can do it.
- Email first, SMS pointer. Correct for anything with detail: a disease exposure notice, a registration process, a policy change. The SMS carries one sentence and a pointer, not a summary.
- All three at once. Four scenarios only. Using it more often is not caution, it is habituation, and it costs you the response when it matters.
Channels Kastr does not have, so this page is not misleading you. There are no push notifications — Kastr sends SMS, email and voice, and nothing else. There are no attachments and no photo or video sharing, so an email newsletter with a PDF attached is not something to plan around. There is no digest batching: a family that lands in two sends in one afternoon receives two messages, which is a reason to be deliberate about cadence rather than to rely on the platform for restraint. And audience resolution covers specific people and everyone in the district, with no grade, school or route presets.
Questions people actually ask
Should a school text parents or call them?
Text for speed and reach; call when the message needs tone, when the household is on a landline, or when it is a conversation rather than an announcement. For most district-wide operational messages the right pattern is SMS with voice failover on the numbers where the text failed, which costs about a tenth of sending both to everybody.
What does it cost a district to phone every family versus text them?
At illustrative rates, roughly $12 per thousand for a 60-second call against $4 per thousand for a one-segment English text — and around $12 per thousand for an accented Spanish text, which is three segments. The comparison that matters is parallel sending against failover: failover reaches nearly the same households for about a tenth of the additional cost.
Which school messages should go out on every channel at once?
Four: an active safety incident, the all-clear that follows it, a same-day closure, and an unplanned early release. Everything else should pick a primary channel. A district that sends everything everywhere has taught its families that no channel is worth checking.
Is email still worth using for school notifications?
Yes, for a specific job: the long form. Email is the only channel with room for the full letter, the policy, the process and the links, and it is the only one a family can find again in March. It is not a notification channel, and a district relying on email for anything time-critical is not reaching most of its families.
How do translated messages differ across channels?
Text and email are read in the family's language and the quality is the translation's quality. Voice is only as good as the recording, and a synthetic read in a language no one on staff speaks should be checked by someone who does before it reaches twelve thousand households. Reading a translation preview before sending is the cheapest quality control available.
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.