Shortlisting candidates

Shortlisting is not filtering. Filtering removes people who do not qualify, and it can be done from a form. Shortlisting decides who your team spends its most expensive hours on, which is a harder job, and most processes run it at resume-screen speed with criteria invented while reading. Below is a method you can apply this week, and then the two arguments that matter: how long the list should be, and which of your criteria are doing nothing.

The shortlisting method, in six steps

Six steps, in this order. The order carries most of the value, because criteria written after you have started reading are criteria fitted to whoever you happened to see first.

01
Write the disqualifiers before you open the pile

Two or three conditions that make somebody impossible for this role, all of them checkable from the application without a call.
- Location and work authorisation, where they are genuinely fixed rather than preferred
- A licence, clearance or certification the role legally requires
- The compensation band, checked as a range rather than as a number
- Anything requiring judgement is not a disqualifier. It is a scoring criterion and it belongs in step two
Where it fails: Years of experience gets used as a disqualifier because it is the easiest field to sort on. It removes the strongest candidate roughly as often as the weakest one.
02
Define what you are scoring, and what each level looks like

Four to six criteria, drawn from what the role has to settle rather than from the advert.
- For each criterion, write down what a strong answer looks like and what a weak one looks like, before any scoring starts
- Weight them, because an unweighted scorecard silently treats communication and domain depth as equally decisive
- Every criterion must have a source of evidence attached: the application, a call, a reference, or the interview
03
Read for evidence rather than for keywords

The keyword is in every application, because every candidate has read the advert. The evidence is in a small fraction of them.
- Look for the outcome attached to the claim: what changed, over what period, and what this person actually did in it
- Look for scale, because the same job title describes different work at forty people and at four thousand
- Look for the gap between what the resume claims and what the work history could physically have contained
04
Verify the three things the shortlist depends on

None of these are on the resume, and all three change the shortlist more than anything that is.
- Availability: what notice they actually have to serve, and what is genuinely negotiable
- Compensation expectation against your band, established before your panel invests an hour
- Interest in this role specifically, which is a different question from interest in a new job
Why this step exists: A shortlist that has not been spoken to is a list of documents. Every profile we put in front of a client has been called and verified, because these three answers move the list more than any amount of re-reading does.
05
Rank, then cut to what your decision-makers can actually meet
- Set the number from the panel capacity you really have, rather than from a sense of how thorough the list should look
- Rank rather than group. A tie means the criteria were not discriminating, and the fix is the criteria rather than a longer list
- Anybody who is on the list because removing them felt harsh is not on the list
06
Write one line per name, then hand it over

This is the artefact that makes an interview useful, and it is the step most often skipped.
- What this person is here for, in one sentence
- The strongest piece of evidence, and where it came from
- The known risk, stated plainly, because a shortlist with no risks listed has not been read carefully
- What the interview should therefore probe, so the hour is spent on the open question
What it changes: A panel handed names spends the hour discovering. A panel handed names with reasoning attached spends it deciding, and that difference is worth more than any additional interview round.

How many candidates should be on a shortlist?

The common answer is eight to twelve for every hundred applications.

“If you have 100 applicants, shortlisting 8 to 12 of them is a feasible goal.”

4 Corner Resources

That is an answer about the screener's confidence rather than about your panel's capacity. A longer shortlist is what you produce when the screening was not conclusive enough to defend a shorter one, and the entire cost of that hedge is paid by the people who have to sit in the interviews.

The counter-argument is in the numbers further down this page, and in the workings behind them.

Criteria that look objective and are not

Each of these sorts reliably, which is why it survives. Sorting reliably and predicting well are different properties.

Years of experience

A proxy for exposure. Exposure at a company where the questions arrived pre-framed and the hard parts were somebody else's job is not the same exposure, and the field cannot tell the two apart.

Current company tier

Screens for who was hired somewhere else, against somebody else's bar, for somebody else's role. It is the most efficient available method of arriving at the same shortlist as your competitors.

Notice period

A real constraint that becomes a poor criterion when applied before anybody has asked what is negotiable. Buyouts, early releases and unused leave are common enough that filtering here removes strong people over a fact that was not final.

Job stability

Two short stints read as a pattern and are usually two ordinary events. Applied as a filter it disproportionately removes people who left companies that were badly run, which is not the trait you were screening for.

Education

The most defensible-looking line on a scorecard and, outside regulated roles, the one least connected to what the role has to settle.

What to hand the panel with the shortlist

The same brief every interviewer has read, rather than a link to the advert.
One line per candidate on why they are on the list.
The stated risk for each candidate, with the question that would resolve it.
The scorecard as already filled in for the completed stage, so nobody re-asks what has been answered.
Nothing that would let a panel member rank on a criterion you did not choose.

If you want a structured way to compare the people already on your list, the candidate fitness indicator scores them against the role rather than against each other.

What a list that is too long actually costs

A longer shortlist does not increase the chance of finding the right person. It increases the number of comparisons, and comparison fatigue reliably ends with the most agreeable candidate rather than the most capable one.

Every additional name is an hour from somebody senior enough to be expensive, and those hours come out of the work the role was opened to support.
The bar drops across a long day of interviews, so candidates late in the sequence are assessed by a tired version of your process.
The gap between the first and last interview widens, and the strongest candidate closes elsewhere inside it.
A padded list also teaches the people producing it that padding is acceptable, which is how a screening standard erodes over a year.

The other side of the same decision: what it costs when the wrong name comes off the list.

How Continuity1 runs this funnel

The screening above is the job. These are the numbers it produces when a function owns it end to end, set against the published benchmarks for the same market.

Read the first two together. A shorter list is only defensible if the people on it convert, and that is the pair that decides whether shortlist length was the right lever.

2 per 100: Profiles reviewed that reach your shortlist; The industry rule of thumb is 8 to 12. We send 2, and the filtering never lands on your calendar.; 4 Corner Resources
1 in 3: Shortlisted candidates you meet who become the hire; Aligned engagements run nearer 1 in 2, distant ones nearer 1 in 10. The market takes about 180 applicants to make one hire, and that sifting lands on your team rather than ours.; Continuity1 tracked engagements
~3: Interviews your team sits in, per hire; Ashby puts technical roles at 17.6 interviews per hire across the whole process, up 52% since 2021. The rest of that load sits with the function, not with you.; Ashby talent-trends report
95%: Offers that close inside your stated band; 20 of the last 21. A flat fee earns nothing from an inflated offer; a percentage of CTC earns more.; Continuity1 tracked engagements

Every brief becomes a success profile before sourcing starts, calibrated with the people who will manage the role. That calibration is the step most hiring skips, and it is why a shortlist either matches the job or matches the job advert.

You review a scored shortlist and make the calls. The filtering never lands on your calendar.

Share a role See the workings and the citations

Questions teams ask

What is the difference between screening and shortlisting?

Screening removes people who do not qualify and can be described in rules. Shortlisting ranks the people who do and cannot, because ranking requires a view on what the role has to settle. Processes that treat them as one activity end up applying rules to a judgement problem, which is how a shortlist ends up sorted by years of experience.

Who should build the shortlist, the recruiter or the hiring manager?

The function builds it, the accountable owner sets the bar it is built against, and the two should be calibrated before sourcing rather than argued about after the first list. A hiring manager building their own shortlist is doing the job because nobody else could apply the bar, and the role stays open while they read.

Can shortlisting be automated?

The disqualifier step can, and should, because it is rules against fields. Steps three to six cannot, because each one depends on evidence that is not in the application: what the person actually did inside the claim, what they will accept, and what they are genuinely interested in. Automating those steps produces a faster version of the same wrong list.

Should we tell candidates they were not shortlisted?

Yes, and treat it as an operational decision rather than a courtesy one. Silence at this stage is the single most common complaint about hiring processes, it is the cheapest thing on this page to fix, and the people you leave in silence are in the same market you will be sourcing from next quarter.

How do you shortlist when every application looks the same?

That is usually a brief problem rather than a reading problem. When an advert asks for a list of tools, every application will contain the list of tools, and no amount of re-reading will separate them. Rewriting the brief around what the role has to settle changes what candidates write about, which changes what there is to read.