How Working Together is scored: the 1–7 scale explained
Behaviour statements are scored on a 1–7 scale, and Working Together has its own logic at each band. This guide shows what assessors can see at scores 4, 5, 6 and 7, and how moderation keeps marking consistent.
The scale and what it measures
Assessors score your statement against the Working Together descriptor for the grade, using the Civil Service's 1–7 scale, where 1–3 fall short in different ways, 4 is acceptable, 5 is good, 6 is strong and 7 is excellent. Departments set their own pass marks, so check the job advert or invitation email for what you need; as of July 2026, many campaigns expect 4 or above at sift, but this varies.
For this behaviour, the score reflects three things: the depth of your personal actions, the presence of the behaviour's core threads (trust, inclusion, constructive conflict) and the strength of the result. A statement can be well written and still score 3 if the evidence is thin.
What each score looks like
| Score | What the statement shows |
|---|---|
| 3 | A teamwork story with vague actions, no real disagreement handled and a result with no number |
| 4 | A genuine example with clear personal actions and an outcome, but one thread is thin: conflict glossed over, others' views decorative, or nothing that lasted |
| 5 | Real friction constructively handled, different views visibly shaping the outcome, a quantified result benefiting more than one party |
| 6 | All of that, plus scope: collaboration across boundaries, influence without authority, and a mechanism or relationship that outlasted the task |
| 7 | Compelling evidence at organisational scale: competing senior interests aligned, trust built deliberately, durable change, results with strategic weight |
Notice that the jump from 4 to 5 is about depth, while the jump from 5 to 6 is about scope and durability. This is why grade matters: the scope expected for a 5 at SEO differs from a 5 at AO.
How moderation works
Sifts are not one person's opinion. Panels of trained assessors score statements, usually independently, and moderation processes exist to keep marking consistent: sample checks, second marking of borderline statements and calibration against agreed standards. The details vary by department and campaign, but the principle is the same — your score reflects the descriptor, not one reader's taste.
Two practical consequences follow. First, write for a stranger: the assessor knows nothing about you, your employer or how impressive the context was, so the statement must carry the evidence on its own. Second, do not chase style. Plain, specific sentences with concrete actions survive moderation far better than polished abstractions.
For example, 'I built strong relationships across the department' reads well and scores poorly. 'I met each of the four team leaders separately before bringing them together, and used six weeks of call data to settle the rota dispute' scores, because a moderator can point to the actions.
Using the scale to improve your draft
Read your draft and score it honestly against the table. If the disagreement is missing, put it back in. If the result has no number, find one: how many people, how much time saved, what percentage changed, what continued afterwards. If you are applying at SEO or above, ask whether anything outlasted the task. Small, honest additions like these are what move statements from 4 to 5 and beyond.
Tip
Aim your statement at 5 before you aim it at 7. Depth first — real conflict, real inclusion, a real number — then scope. A statement with depth at the right grade level always beats a grand story told thinly.
Frequently asked questions
It varies by department and campaign. Many expect 4 or above on each behaviour, but some set higher pass marks or rank candidates when places are limited. Check the job advert and any sift guidance in your application account, because the pass mark is set per campaign.
A single weak sentence rarely decides a score, but a missing element can. Assessors score the evidence as a whole: if your statement never shows disagreement handled or never gives a countable result, the statement as a whole lacks that thread. Fix gaps rather than polishing sentences.
Yes. They score against the descriptor for the grade you applied for, which is why pitching matters. An excellent AO-level story in an SEO application scores against SEO expectations and will likely fall short on scope, however well it is written.
No, but 7 is deliberately rare. It is meant for compelling, complete evidence at the right scope. Most strong candidates score 5 or 6. Rather than chasing 7, make sure you are not leaving a 5 on the table through vague actions or a result without numbers.
Next step
Get the free STAR template — the same structure used in every example on this site.
Download the template pack