Workflow

From corridor concern to support plan: a workflow that survives a busy term

Most SEMH support starts with a passing comment from a TA. Whether it becomes support at all depends on what your school does in the following 48 hours.

The Phoenix Ember team · 23 August 2026 · 6 min read

Here is where most SEMH support actually starts. Not with a referral form. With a teaching assistant catching a SENCO's eye at the end of break: have you noticed anything with Sapna lately?

That sentence is the most valuable early-warning signal a school has. Frontline staff notice change weeks before it reaches any formal system. And in most schools, that sentence has roughly a 48-hour half-life. If nothing captures it, it is displaced by the next hundred urgent things, and it resurfaces months later as a bigger problem with a longer recovery.

The difference between schools that catch children early and schools that are repeatedly surprised is rarely commitment. It is whether a concern has somewhere to go.

Why concerns die

Watch a concern fail to become support and you will almost always see one of three breaks:

  • No route. The member of staff did not know where to put it, so it stayed verbal. Verbal concerns are lost concerns.
  • No owner. It was written down, in an email or a corridor conversation relayed second-hand, but it landed on nobody in particular. Things addressed to everyone are actioned by no one.
  • No review date. Support was actually put in place, and then nobody came back to ask whether it worked. The plan ran on unexamined, or quietly stopped, and the record ends mid-sentence.

Notice that none of these is a knowledge problem. They are plumbing problems, and plumbing can be fixed.

The minimal workflow

The workflow that survives contact with a real term has five stages, and none of them is sophisticated.

Capture, in the moment. The concern goes into one agreed place, in the words of the person who noticed, dated, within minutes of being noticed. The bar for raising must be low: a worry, not a proven problem. If staff fear being wrong, they will wait for certainty, and certainty arrives late.

Route, to a named person. Every concern lands with exactly one owner, usually the SENCO or a pastoral lead, whose job at this stage is triage, not solution. The person who raised it should be able to see that it landed. Silence teaches staff to stop raising things.

Decide, visibly. Triage has three honest outcomes: act now, watch deliberately with a date attached, or reassure and record why. All three are legitimate. What is not legitimate is the fourth, unofficial outcome: pending forever. A watching decision without a date is a pending decision.

Act, with a baseline. If support starts, score the child first, however your school measures. Thirty seconds of baseline turns everything that follows from anecdote into evidence, and the case for that is the whole of our guide to evidencing provision.

Review, on a date that was set at the start. The review date is part of the plan, not a thing someone remembers later. At review, the question is plain: better, worse, or the same, compared with the baseline, and what do we do accordingly.

Make the system remember, so people can forget

The five stages fail in schools that run them on goodwill, because goodwill is exactly what the busiest weeks consume. The fix is to move memory out of heads. Review dates that surface themselves. Concerns that visibly age until someone acts. A record that shows, at a glance, every open loop for a child. Whether you build that with a spreadsheet and fierce discipline or with a system designed for it, the property you are buying is the same: no child's support depends on a busy adult remembering.

The September test

There is a second reason to run concerns through a system rather than through people, and it arrives every September. Staff leave, classes are re-formed, and the child starts the year with an adult who was not there for the story so far. In a school with a working concern record, the handover exists automatically: here is what was noticed, tried, and learned. In a school without one, the new teacher starts from zero, and the child pays for the repetition.

A concern raised in the corridor deserves better than the corridor's memory. Give it a route, an owner, a baseline and a date, and the rest of SEMH provision becomes very much easier to do and to prove.

Phoenix Ember is a wellbeing and SEND tracker for schools: assessment, interventions, alerts and evidence in one place.

See how it works