If you are the one accountable for it
See the signal. Not the person.
HABLA™ helps leadership see which topics are repeatedly finding their way to the surface, what changed against a comparable period — the same weeks last year, and last quarter — and what leadership decided to address, who owns it, and where it stands. Together they turn what surfaces into something a leadership team can put on an agenda.
It does not tell leadership who tapped them, how many unique people they represent, or whether a person has a diagnosis.
What arrives on the first screen
- What's Surfacing
- Emerging Movement
- Possible Resource Gaps
- Action Center
- Benchmark
- Methodology
Every plan includes Aggregate monthly reporting. What each plan includes
What arrives, and in what order.
The first screen answers six questions in the order a leadership team actually asks them, and every section on it is constrained by the two floors below. What follows is the whole of it, written down so it can be argued with before you sign rather than discovered after.
Every section on that screen releases whole or not at all. A breakdown is published only when every one of its rows is either at or above the reporting floor or at zero — one row between the two suppresses that breakdown for the period, rather than printing it with a gap where the small row was, because a gap can be subtracted back. So a period can report nothing, and the report says so rather than quietly printing something else in its place. What holds a section back is the population floor below: where publishing a wider figure beside a narrower one would let a unit too small to report on its own be worked out by subtraction, the wider figure is held back for that period as well. The floor does not move to make a section appear.
What's Surfacing
Every topic in your library, largest first — or, for that period, nothing at all. It releases whole or not at all: every topic has to be at or above the reporting floor, or at zero, and one topic between the two suppresses the section rather than leaving a gap in the list.
Emerging Movement
What changed against a comparable period — the same weeks last year, and last quarter.
Possible Resource Gaps
Where people may be having trouble using support that already exists.
Action Center
What leadership decided to address, who owns it, and where it stands.
Benchmark
HABLA compares this deployment against privacy-safe sector and national aggregates, with every row carrying its denominator. Two organizations can differ because their deployments differ in age, awareness or adoption — not because anything about the people differs. No cross-organization figure has been published yet, and none will be until one clears the contributor floors this page sets out.
Methodology
How to read the report. Permanently visible, never collapsed by default, never a footnote.
One signal, in the five parts it arrives in.
“HABLA cannot conclude” is a required field, not a disclaimer. It ships beside every signal in the same visual weight as the interpretation, because a suggested response printed without its limits is exactly how a count becomes a conclusion.
- Signal
- “I don’t know who I’m supposed to ask” increased from the previous valid reporting period.
- Why leadership should care
- People may be encountering friction understanding where support resources exist.
- HABLA cannot conclude
- This does not establish how many people are experiencing the issue, or why the increase occurred.
- Possible organizational responses
- Simplify the resource pathway · Introduce named support contacts · Publish bilingual resource information · Review how resources are explained at onboarding.
- Action
- Leadership records a decision, an owner and a date — three of the fields in the organization’s own decision record. Every response HABLA offers is organizational — it never suggests approaching a named person, team lead, or manager.
The drawing below is the model HABLA publishes for what happens next — signal, insight, strategy, action, learn — and the band across it is where HABLA stops and a person decides. Insight and strategy will be the paid add-on; signal, action and learn are part of every plan.
Signal to learning
A pattern reaches the surface, HABLA reads it and offers possible responses, a person decides and acts, and HABLA keeps watching the signal afterwards without claiming the action caused what followed. This is the model HABLA is building towards: none of the five stages is running for a customer today.
-
A topic that keeps reaching the surface across an organization, as aggregate activity that has already cleared both reporting floors. Core, in every plan.
-
Trend, persistence, timing and context, read without identifying anyone. The deterministic view of a topic is core; the generated reading of it is the paid add-on.
-
Possible organizational responses, or ways a practice that appears to be going well could be tried somewhere else. Options a person chooses between — never an instruction, and never a suggestion to approach a named person. The paid add-on.
-
Everything on the far side of this band is done by people in the organization. HABLA produces the signal, the reading of it and the options; it does not decide, does not act, and does not contact anyone.
-
What leadership decides to do, and when it will be reviewed. The organization acts; HABLA never contacts anyone or intervenes for them. The organization’s own decision record keeps what was decided and done; the decision, the administrator who owns it and the date are three of its fields. The record holds no audience-member identity; what prompted it was a pattern in anonymous activity, and it never claims an action caused an outcome. Core, in every plan.
-
Watching does not stop at the decision. HABLA will keep the signal under watch and report what the aggregate activity did in the reporting periods after a recorded action — stated as a comparison of closed periods, and never as a claim that the action caused the change. Core, in every plan.
Movement, not volume.
The largest topic this period is usually the largest topic every period, and a report that only ever restates it does not earn a second opening. The same topic against a comparable period is a different object: it is a question with a date attached, and a date is where a decision can be made.
These are the six questions the reporting engine answers, and the form each answer takes. Every one of them is a question about the organization rather than about a person, which is what makes the answers safe to give.
| A leader asks | The report answers with |
|---|---|
| Is this getting better or worse? | The same period, last year and last quarter, side by side. |
| When does this reach people? | Access timing — hour of day and day of week. That is when the experience was used, not when a need began. |
| Where is it surfacing? | By site and by audience pack — which reflects the placement experience someone used, not where anyone was — subject to both floors below. |
| What is going well? | Appreciation topics sit alongside concern topics, so the product is not only a complaints channel. What is going well is reported in the same shape, and under the same floors, as what is not. It is never scored against a team and never set against a concern to cancel it: a good week and a bad one are two facts, not a balance. |
| Did what we changed do anything? | The action record sits against the periods either side of it — what was decided, by whom, and when. |
| What should we look at first? | Movement, not volume. |
Two independent floors, and no contract changes either.
The activity floor. Below ten activity events, a value is suppressed rather than shown. The floor is applied to each topic in each hour of each day, before the month is added up. So a topic that never reaches ten activity events inside a single hour reports a zero for the month — and that zero means no activity there reached the floor, not that nothing happened. Ten activity events is a count of interactions, not a count of people — HABLA cannot count people, and a suppressed cell means “below the reporting floor of ten activity events”, never “fewer than ten people”.
The population floor. A placement or audience pack is only separately reportable if the population it covers is at least 50 people. Below that, its activity is not shown as its own row — it rolls up into its parent site. If the parent is small too, it rolls up again. An organization of thirty people receives organization-level totals and nothing beneath them, which is the correct outcome rather than a degraded one: the breakdown it would otherwise get would be a breakdown about identifiable people.
Neither is a setting. There is no per-customer threshold field anywhere in the platform, so there is nothing for a large contract to negotiate. Cross-organization figures carry a third floor: one is published only where at least ten organizations contribute, no single organization exceeds 40% of it, and the top three together do not dominate it. No such figure has been published yet, and none will be until one clears all three — the combined limit is a number HABLA has not set, and the code refuses to publish rather than assume one.
What a person's own answer becomes. If someone says how much a statement fits them, that answer arrives here as a distribution — a little,quite a bit, very much, as three shares of reportable activity — with each cell clearing the floor on its own, and if any cell falls below it the breakdown is suppressed whole rather than shown with a gap. A residual that can be recovered by subtraction is not suppressed. It is never averaged, never becomes a severity, and is never attached to a person. The person is told what their answer becomes: one of three shares of the reportable activity for that statement, never a number about them. HABLA counts interactions, not unique people. A share of people is not a figure HABLA can produce, because nothing in it counts people.
Below the floor, nothing is shown
Two floors, not one. A count is suppressed below ten activity events, and the floor is applied to each topic in each hour of each day, before the month is added up; a placement or audience pack is only reportable on its own if it also covers at least fifty people.
-
Suppressed. Not rounded, not estimated, and never shown next to another number that could reveal it by subtraction.
-
A fixed threshold. It is not lowered for a larger contract — that refusal is the product.
-
Clears the activity floor. Reported as forty-two activity events — never as forty-two people.
-
The second floor. A placement or audience pack covering fewer than fifty people is not reportable on its own however high its count climbs — its activity rolls into the parent site.
What leadership never receives
- User profiles
- Individual risk scores
- Anonymous narrative files
- Session history
- Raw anonymous exports
- A “high-risk employee/student” list
- Per-person engagement dashboards
There is also no event list — no row reading “Money and bills · 3 November · 15:47”. In a small unit, one timestamped event plus ordinary knowledge of who was working that afternoon identifies a person, and no amount of aggregation elsewhere in the report undoes it.
HABLA is also not a route into clinical care. It offers language, resources, and — if a person chooses it — a way to reach someone they named themselves. It does not refer anyone, book anything, notify a clinician, or escalate, and no configuration adds that.
There is also no live view and no alert. HABLA is not monitored in real time, no one is notified when a topic is selected, and nothing in the product tells you that someone is in difficulty now. If you need someone watched in real time, HABLA is the wrong product, and the page written for the person says the same thing from the other side.
HABLA will not rank organizations by risk, distress, or wellbeing. “Largest change in reportable activity” is the factual form that comparison takes. “Highest-risk schools” is not a view that exists, and no configuration produces it.
The panel every report carries
This is the required wording, in full. It is not optional, not collapsible by default, and not a footnote — it ships with every report surface, export, and board brief, in the same words each time.
How to read this report
HABLA counts interactions, not unique people. Counts do not measure prevalence, diagnosis, or severity. A site value reflects the organizational placement experience used and does not confirm the person's physical location. An audience-pack value reflects the experience accessed and does not authenticate the visitor as a member of that group. Differences may reflect population size, awareness, access, placement, repeat use, or other factors. Use these signals to identify questions worth examining — not to identify individuals or estimate how many people have a condition.
Read this next to what the person meets.
The two pages describe one system from its two ends. Everything on this page is what remains after the other page’s guarantees are met — which is the only order in which either of them means anything.