The families who never respond: eleven causes, ranked, and what to do about each
Every school has a list of families nobody can reach, and the list is treated as a fact about those families. It is usually a fact about the data. Here is how to tell the difference, in order of how often each cause turns out to be the real one.
| # | Cause | Type | The diagnostic | The fix |
|---|---|---|---|---|
| 1 | No contact point on file at all | Data | Query students with zero contact points; it is never zero students | Front-office capture at any contact, plus a targeted call list |
| 2 | Number invalid or reassigned | Data | Validation state, plus repeated terminal failures | Correct at source, and record where the correction came from |
| 3 | Wrong guardian ranked first | Data | Ask the parent who called which adult usually gets school messages | Guardian rank on the student record, not a workaround list |
| 4 | Language mismatch | Data | Compare household language against the language each send went out in | Record preferred language properly; translate before sending |
| 5 | Carrier filtering of unregistered sender | Technical | An entire carrier's worth of recipients silent while others respond | 10DLC brand and campaign registration |
| 6 | Consent never recorded for the channel | Data | Consent state distinct from validation state | Capture consent explicitly; do not infer it from a number existing |
| 7 | Email lands in spam or promotions | Technical | Email opens near zero for a subset while SMS works | Sender authentication, and stop sending marketing-shaped email |
| 8 | Message volume has trained them to ignore you | Behavioural | Engagement decays across a term while data stays clean | Fewer, better messages; a district cadence rule |
| 9 | Messages arrive at unusable times | Behavioural | Send timestamps against shift-work patterns in your community | Quiet hours, and stop sending at 4:55pm on Fridays |
| 10 | No reply path, so silence is rational | Design | Ask families whether they know they can reply | A monitored reply channel, or explicit "do not reply" plus a number |
| 11 | Genuine disengagement | Human | Everything above has been excluded, in writing | A person, not a platform |
Why the ranking matters more than the list
Most districts start at the bottom of that table. The conversation opens with disengagement, moves to message design, and never reaches the top three, which are boring, unglamorous data problems that account for the majority of cases.
The practical consequence is misallocated effort. A district will run a family-engagement campaign for the families who cannot receive the campaign. The cheapest large win available to almost every district is closing cause one: the students with no usable contact point on file. That is typically 3–8% of enrolment, it is a finite list, and it can be worked through by the front office in a term.
Work top-down. Do not touch causes eight to eleven until one to seven are excluded for the family in question, because the fixes are expensive and the diagnosis is usually wrong.
A termly contact-data routine that is not a data project
Three hours per school per term, run by the front office rather than by IT:
- Hour one — the zero list. Produce the list of students with no contact point of any kind. Call the ones with a number in the SIS but nothing in the messaging platform first, because those are usually a sync problem rather than a family problem, and they resolve in bulk.
- Hour two — the failure list. Every contact point with a terminal failure since the last review. Correct or retire each one. A number that has failed three times is not going to start working.
- Hour three — the rank and language pass. For families who have contacted the office this term, confirm which adult should be first and which language they want. This is the highest-quality data you will get all year and it arrives free, through the counter.
The honest note about tooling. Kastr does not detect stale contact data automatically. There is no engine that raises data-health issues for you — contact points carry validation and consent state, and roster syncs record failures, but nothing proactively tells you a family has gone dark. If a vendor claims automated data-health detection, ask them to demonstrate an issue being raised during the demo, not one being cleared. This is a human routine, which is why we have written the routine out rather than describing a feature.
Adding channels: when it helps and when it is noise
The instinct after a failed campaign is to add a channel. Sometimes that is right and often it is not.
It helps when the failure is channel-specific: a household with no data connection reached by voice instead of email, or a family whose email bounces reached by SMS. Kastr fails SMS over to voice on terminal failure specifically for this case, and links the voice attempt back to the original send so the record shows what was tried.
It is noise when the same message goes out on three channels to the same family every time. That trains people to ignore all three, and it is expensive: voice is the costly channel and sending it as a default rather than as a fallback is how a telephony budget disappears by March.
The rule that holds up: one channel per message by default, chosen by urgency; failover on delivery failure rather than in parallel; all channels simultaneously only for genuine emergencies, where the redundancy is the point and the cost is irrelevant.
The escalation ladder, and where technology stops
- Retry on the same channel, once, at a different time of day.
- Switch channel. If SMS fails, voice. If email is silent, SMS.
- Try the second guardian. Frequently the actual fix, and frequently skipped.
- Try the emergency contact. Often a grandparent who sees the child daily.
- Send paper home with the student. Slow, and it works more often than anyone expects.
- A person telephones, from a number that shows as the school, at a time working parents can answer.
- A home visit, where your district has staff who do this.
Steps one and two are automatable. Step three should be. Steps four to seven are people, and no communications platform changes that — including ours. What a platform can do is make sure the twenty families who need step six are a known list rather than a discovery, so the human effort goes where it is needed instead of being spread across everyone.
Questions people actually ask
How do we tell a bad phone number from an ignored message?
Delivery state. A terminal failure means the number did not accept the message and is a data problem; a successful delivery followed by no response is a behaviour or design problem. The trap is a message that was accepted by a carrier and then filtered, which looks like success and reached nobody — that is why the two states must be reported separately.
What percentage of families are typically unreachable?
In districts that measure it honestly, 3–8% of students have no usable contact point at all, and a further slice is reachable only on a channel nobody checks. Any vendor quoting you a figure above 95% reachable is almost certainly counting contact points rather than students. We publish no such figure for exactly that reason.
Does adding more channels help or just add noise?
It helps as failover and hurts as a default. Sending every message on every channel trains families to ignore all of them and consumes the telephony budget, because voice is expensive. Use one channel chosen by urgency, fail over on delivery failure, and use all channels at once only for genuine emergencies.
Which guardian should be first on the contact list?
The one who handles school logistics, which is not always the one listed first in the SIS and is frequently the cause of "we were never told" in separated households. Ask at the counter, record it as guardian rank on the student record, and do not maintain a parallel list — parallel lists go stale within a term.
When should we stop messaging and go to the door?
When channel switching, the second guardian, the emergency contact and paper home have all been tried and recorded. At that point continued messaging is not outreach, it is documentation. The value of working the list top-down is that it makes this a short, specific list of families rather than a vague sense that some people are hard to reach.
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.