Why Civil Service Applications Fail at Sift

How does a sifter read your application?

A sifter works from a marking scheme built from the advert. The way answers are scored, the pass mark and who does the sifting all vary by campaign, so do not assume what worked in one campaign applies to the next.

Some things hold steady. Sifting is name-blind, so only your evidence counts. Only behaviours, technical requirements and experience are sifted. Strengths are not named in adverts and are assessed at interview. Vague statements are hard to treat as evidence, and no evidence means no score.

Is your evidence pitched at the wrong grade?

This is the most common way capable applicants lose marks. Behaviours are scored against the descriptor for the advertised grade, not the behaviour name. At EO, Changing and Improving asks you to review your own and your team's work and suggest improvements. At HEO it asks you to work with others to identify improvements, simplify processes and make it safe to challenge. An HEO applicant whose example stays inside their own desk reads, to a sifter, as EO evidence.

Grade shows through scope, complexity, decision-making and influence, not bigger words. If your best example genuinely sits a grade below, look for another where your role stretched beyond its title. Never inflate what you did.

As an example, a weaker version reads "I suggested a tidier folder structure for our shared drive and it was adopted." A version pitched at HEO reads "I reviewed how three teams logged casework, agreed one template with their managers and trained 12 colleagues to use it, which cut duplicate entries by a third."

Do you describe duties, use "we" or leave out the result?

"I was responsible for" tells a sifter your job title, not how well you did it. They are looking for the action you took and what changed. "We" has the same problem, because a team result gives them nothing to score against you. A claim with no outcome reads as assertion.

An example of a duty statement reads "I was responsible for processing grant applications and answering queries." A stronger version reads "I cleared a backlog of 140 grant applications by sorting them by risk and agreeing a standard reply for the five most common queries, which cut average response time from ten days to four."

An example of a "we" statement reads "We redesigned the intake form and errors went down." A stronger version reads "I reviewed 60 returned forms, drafted a shorter intake form, tested it with two colleagues and won my manager's approval to adopt it. Errors fell by about a quarter in three months."

If you have no number, describe what observably changed. Never invent a figure.

Have you answered what the advert actually asked?

Sifters mark line by line against the criteria and behaviours named in the advert. A statement that spends most of its space on career history and never touches Delivering at Pace, for instance, leaves that behaviour with nothing to score. The fix is to list every essential criterion and behaviour first, then give each its own example, using recognisable wording where it is true to what you did.

Format matters too. If the advert asks for a set length or separate behaviour answers, give exactly that. Text past a limit may be cut off or not read, so your strongest point should never sit in the last sentence.

Eligibility is the quiet one. Check the advert for who can apply, right to work, security clearance, location and any required qualification. An application that does not meet a stated requirement can fall before your evidence is scored at all.

Are your statements vague, or is every example the same?

Generic lines such as "I have strong communication skills and build good relationships" could describe anyone, so a sifter cannot treat them as evidence. An example of a specific version reads "I wrote a two-page briefing explaining a funding change to 30 local providers and answered their questions at two webinars, after which no provider needed escalating."

Repeating one example across several behaviours is a related trap. If Leadership, Working Together and Delivering at Pace all rest on the same project, the sifter sees one piece of evidence, thinly spread. Choose a different example for each. Where one project truly spans two criteria, make the action for each clearly different.

What should you check before you submit?

• Each essential criterion and named behaviour has its own example.
• Each example is pitched at the advertised grade's scope and decision-making.
• Every example says what you did, in the first person, not what the role involves.
• Each one ends with a result, a number where you can defend it or an observable change where you cannot.
• You have not reused one example across criteria without different actions.
• You are inside the word limit and have used the format the advert asked for.
• You meet every stated eligibility requirement.
• Everything you claim is true and you could explain it in detail at interview.

Where Koda fits

Koda helps you apply this to your own experience. It works from your real evidence and matches it to the role and grade you are applying for, with nothing invented, and you stay in control of every word.

Frequently asked questions

Related guides

• How to Write a Civil Service Personal Statement
• The STAR Method for Civil Service Applications (and STARR)
• The 9 Civil Service Behaviours, Explained