We publish whether it worked.
Once a year, four numbers, covering everyone we assessed rather than a flattering slice of them. Including the one number nobody in this sector shows you: how often we said no.
These are what we want to be judged on.
Support hours reduced within 12 months
Average change in weekly hours per child, first year to date.
Median days from first contact to plan
First message received to a signed Independence Plan.
Independence milestones met against projection
Milestones reached, compared with the date we projected at Plan.
Children assessed and referred elsewhere
Assessed children who did not become clients, and where they went.
Each one is uncomfortable in a different way.
Support hours reduced within 12 months
The average change in weekly support hours per child. For a provider paid by the hour this is a revenue figure pointed the wrong way, which is precisely why it belongs at the top of the list.
Median days from first contact to plan
From your first message to a signed Independence Plan. Median rather than average, so a handful of fast cases cannot hide a long tail of families left waiting.
Independence milestones met against projection
Milestones reached, compared with the date we projected when the plan was written. Not milestones reached in total — reached against what we said.
Children assessed and referred elsewhere
The share of assessed children who did not become clients, and where they went instead. An inverted conversion rate, published where our competitors can read it.
How we'll stop ourselves gaming it.
A published metric with no rules around it becomes a marketing number within two years. These are the constraints we are committing to now, while it is still easy to.
Everyone assessed, not everyone enrolled
The denominator is every child we assessed in the year. Excluding families who left early would make all four numbers look better and all four numbers meaningless.
Published even when it's bad
If a year goes badly, that year is published looking bad. A report that only appears in good years is an advertisement.
The method is published with the numbers
How each figure was calculated, including any exclusion and the reason for it, so a change of method next year is visible rather than convenient.
No child is identifiable
Aggregate only. Where a group is small enough that a figure could identify a family, the figure is withheld and the reason is stated.
You don't have to wait a year to hold us to something.
The annual report is the public version. Your own version arrives far sooner and in more detail: session data in your file within the week, a written review every term, and support hours plotted against the projection in the plan.
If our hours are not coming down and the plan said they would, that is visible to you at the same moment it is visible to us. It is the whole reason the file is yours rather than ours.
- What proportion of children you assess do not become your clients?
- What happened to your average support hours per child last year?
- Can I see a plan you wrote that recommended someone else?
- Who signs off a recommendation, and are they paid on it?
Including us. If we ever stop being able to answer these, that is the moment to take your file elsewhere.
Tell us what you've noticed.
Twenty minutes on the phone with someone who does this every day. No assessment, no cost, and no obligation to become a client at the end of it.
If we're not the right people, we'll tell you who is. That includes naming providers we don't own, and it costs you nothing either way.