# How to shortlist candidates: a practical guide with a worked example

> A shortlist decides where your interview time goes. How to size it, who is eligible, how to rank within it, how to check the near misses, and how to write it up so someone else can check your reasoning.

Sara at SortTheStack, October 19, 2026. Original: https://sortthestack.com/blog/how-to-shortlist-candidates

**To shortlist candidates, decide how many people you can actually interview,
take everyone who clearly meets your must-haves, put them in order by the
evidence for your nice-to-haves, and write down a reason for every name on the
list and every near miss that isn't.** A shortlist is a decision about where
to spend interview time. It isn't a verdict on who is good enough, and treating
it that way makes it both faster and fairer.

This guide picks up where
[screening a large pile of resumes](https://sortthestack.com/blog/how-to-screen-a-pile-of-resumes)
leaves off. You've read every application against the same requirements. Now
you have to decide who gets a call.

## What a shortlist is for

Interviews cost time: preparing, an hour or so each, writing up, and
coordinating calendars. You can't interview everyone who could do the job, so
you choose. The risk is in how you choose.

Two shortcuts are common. The first is shortlisting by impression: the
applications that read best, or the first good ones you found. The second is
shortlisting by a single easy signal, such as years of experience or a
well-known previous employer. Both select the people who are easiest to
evaluate, not necessarily the best. Years of job experience, for example,
correlated just .07 with later performance in a
[2022 re-analysis of selection research](https://sortthestack.com/blog/what-predicts-job-performance).

## Step 1: Set the size before you look

Work backwards from your interview capacity. If you can run six first
interviews next week, and you expect one or two people to withdraw or accept
another offer, a shortlist of six to eight is about right.

There is no research-backed ideal shortlist size. Capacity is the honest
constraint, and setting the number first stops it drifting to fit whoever you
happen to like. If far more people qualify than you can see, add a short step
that tells you more, such as a 15-minute phone screen or a
[short work sample](https://sortthestack.com/blog/how-to-hire-the-best-candidate), rather than cutting
the group arbitrarily.

## Step 2: Start from the must-haves, not the best resume

Sort every screened application into three groups:

| Group                      | What it means                                            | What to do                           |
| -------------------------- | -------------------------------------------------------- | ------------------------------------ |
| Meets every must-have      | Clear evidence for each must-have in the application     | Eligible for the shortlist           |
| Unclear on one or more     | Nothing contradicts a must-have, but nothing confirms it | Ask, before deciding                 |
| Clearly misses a must-have | Evidence shows a must-have isn't met                     | Not shortlisted; reason written down |

Only the first group goes straight to ranking. The middle group is where
qualified people get lost. An application that doesn't mention a requirement
hasn't shown that the person lacks it, and a one-line question by email often
settles it.

This matters at scale. In a survey of 2,275 employers in the United States, the
United Kingdom and Germany,
[Harvard Business School and Accenture (2021)](https://www.hbs.edu/ris/Publication%20Files/hiddenworkers09032021_Fuller_white_paper_33a2047f-41dd-47b1-9a8d-bd08cf3bfa94.pdf)
found that 88% of employers who used recruiting software to filter or rank
applicants said it screened out qualified high-skills candidates at least some
of the time, because they didn't match the exact criteria in the job
description. A person reading quickly does the same thing.

If you don't have a written list of must-haves yet, start with
[must-have vs nice-to-have](https://sortthestack.com/blog/must-have-vs-nice-to-have).

## Step 3: Rank within the eligible group

Everyone in the first group can do the essentials, so rank them on what is left:

1. **Nice-to-haves with evidence.** Count how many each person shows, using the
   same list for everyone.
2. **Strength of the evidence.** "Ran the month-end close for a 40-person
   company" is stronger than "involved in month-end activities". Our guide to
   [what counts as evidence](https://sortthestack.com/blog/evidence-a-candidate-can-do-the-job) has a
   simple ladder for this.
3. **Ties stay tied.** If two people are level at the cut-off, take both, or
   take neither and look again. Don't break a tie on something you didn't
   decide in advance.

## A worked example

Larkspur Partners, a fictional accounting firm used throughout our guides, is
hiring a Senior Accountant. Its two must-haves are a CPA designation and having
run a month-end close independently. Its nice-to-haves are audit working
papers, cloud accounting and a public accounting background. The partner can
run six first interviews.

The figures below are invented for illustration. Of 180 applications, 31 met
both must-haves, 14 were unclear on one, and the rest clearly missed at least
one. Here are the applicants around the cut-off after ranking the 31:

| Applicant | Must-haves | Audit papers | Cloud ledger | Public accounting | Evidence strength           | Decision                   |
| --------- | ---------- | ------------ | ------------ | ----------------- | --------------------------- | -------------------------- |
| A         | Both       | Yes          | Yes          | Yes               | Owned the close and audit   | Shortlist                  |
| B         | Both       | Yes          | Yes          | No                | Owned the close             | Shortlist                  |
| C         | Both       | Yes          | No           | Yes               | Owned the close             | Shortlist                  |
| F         | Both       | No           | Yes          | Yes               | Led the close for a quarter | Shortlist (6th)            |
| G         | Both       | No           | Yes          | Yes               | Led the close for a quarter | Tie with F: shortlist both |
| H         | Both       | No           | No           | Yes               | Owned the close             | Near miss                  |

Applicants D and E (not shown) also made the list. F and G were level on
everything the firm had decided mattered, so the partner interviewed seven
rather than inventing a tie-breaker. Meanwhile, the 14 unclear applicants were
each sent one question. Three answered in a way that put them among the
strongest eligible applicants, and they were added to the shortlist for a second
round of calls the following week.

## Step 4: Check the near misses

Before you send the list anywhere, read the applications just below the line
once more and ask one question of each: _is the only reason this person is
below the line something we said doesn't matter?_ If yes, swap them in.

Then look at the shortlist as a whole. If everyone on it went to the same two
schools, worked at the same kind of company, or shares a background, ask
whether your requirements produced that or your reading did. Field experiments
show the reading can. In a study that sent more than 83,000 fictitious
applications to jobs at 108 of the largest U.S. employers,
[Kline, Rose and Walters (2022)](https://doi.org/10.1093/qje/qjac024) found
that distinctively Black names reduced the chance of employer contact by 2.1
percentage points compared with distinctively white names, with large
differences between companies. In Toronto, resumes with English-sounding names
received interview requests 40% more often than otherwise identical resumes
with Chinese, Indian or Pakistani names, 16% against 11%
([Oreopoulos, NBER, 2009](https://www.nber.org/digest/sep09/why-do-skilled-immigrants-struggle-labor-market)).

Neither study observed your process, and the gaps varied a great deal between
employers. The practical point is that a written list of evidence, applied the
same way to everyone, gives you something to check your instincts against.

## Step 5: Write one line per decision

For each person on the shortlist, write the evidence that put them there. For
each near miss, write why they aren't on it. One line each:

- "B: CPA; ran the close for a 60-person distributor for three years; prepared
  audit files; uses a cloud ledger."
- "H: meets both must-haves; no nice-to-haves shown. Next in line if anyone
  withdraws."

This takes ten minutes. It lets someone else check your reasoning, it keeps the
second round honest, and it's the record you'll want when you look back at the
hire.

## Step 6: Share it in a form someone else can check

If a hiring manager, partner or client is choosing who to interview from your
list, send them the requirements, each candidate's evidence against each one,
anything still unclear, and the near misses. A list of names with "strong fit"
next to each one asks them to trust you. A list with evidence lets them agree or
disagree for a reason. Our guide to
[writing a candidate recommendation](https://sortthestack.com/blog/how-to-write-a-candidate-recommendation)
shows the format.

## When the shortlist is too short

If fewer people meet your must-haves than you have interview slots, don't lower
the bar quietly by shortlisting people who miss one. Look upstream instead:

- **Are the must-haves real?** A requirement that is really a proxy, such as a
  degree or a year count, removes people who could do the job.
- **Did the posting attract the right people?** See
  [how to write a job description](https://sortthestack.com/blog/how-to-write-a-job-description).
- **Is the problem the market or the process?**
  [Not finding qualified candidates?](https://sortthestack.com/blog/finding-qualified-candidates) has
  five things to check.

## Where SortTheStack fits

SortTheStack reads every application against your Must have and Nice to have
requirements and marks each one met, not met or unclear, with the evidence
quoted from the application. Applicants are grouped as Recommended, Possible,
Needs review or Requirements not met, so the unclear ones are visible rather
than silently dropped. You decide who moves to Shortlisted, compare finalists
side by side, and share the shortlist with a hiring manager or client by link.
[See how the screening works](https://sortthestack.com/product).

## Sources

- Fuller, J. B., Raman, M., Sage-Gavin, E., & Hines, K. (2021).
  [_Hidden workers: Untapped talent_](https://www.hbs.edu/ris/Publication%20Files/hiddenworkers09032021_Fuller_white_paper_33a2047f-41dd-47b1-9a8d-bd08cf3bfa94.pdf).
  Harvard Business School Project on Managing the Future of Work and Accenture.
- Kline, P., Rose, E. K., & Walters, C. R. (2022). Systemic discrimination
  among large U.S. employers. _The Quarterly Journal of Economics, 137_(4),
  1963–2036. [doi.org/10.1093/qje/qjac024](https://doi.org/10.1093/qje/qjac024)
- Oreopoulos, P. (2009). _Why do skilled immigrants struggle in the labor
  market? A field experiment with six thousand resumes_ (NBER Working Paper
  15036), as summarised in the
  [NBER Digest, September 2009](https://www.nber.org/digest/sep09/why-do-skilled-immigrants-struggle-labor-market).
- Sackett, P. R., Zhang, C., Berry, C. M., & Lievens, F. (2022). Revisiting
  meta-analytic estimates of validity in personnel selection. _Journal of
  Applied Psychology, 107_(11), 2040–2068.
  [doi.org/10.1037/apl0000994](https://doi.org/10.1037/apl0000994)