Candidate NPS: How to Measure It and What to Do With the Score

Key takeaways

• Candidate NPS measures the percentage of promoters (9 to 10) minus the percentage of detractors (0 to 6) among candidates asked, “How likely are you to recommend applying to our company to a friend or colleague?” It ranges from -100 to +100

• A score greater than zero means more candidates would recommend applying than would advise against it, but that score means less without the candidate’s stage attached. Early rejections typically score far lower than hires, and hires are the most likely to respond to the survey. Assess scores on the furthest stage reached to control for these variables

• Only about 18% of candidates answer the survey in Ashby’s data, so treat candidate NPS as a signal to investigate, never as the only basis for a decision

• Most applicant tracking systems can send an NPS survey; the value comes from breaking the score down by furthest stage, hiring manager, recruiter, and source, next to the hiring data that explains it

What is candidate NPS?

Candidate NPS (Candidate Net Promoter Score, often shortened to cNPS) measures how likely candidates are to recommend applying to your company. It’s based on one question, answered on a scale from 0 to 10. The score is the percentage of promoters minus the percentage of detractors.

The standard wording of the question is: “How likely are you to recommend applying to our company to a friend or colleague?” The question is focused on applying because most of the people you survey were never hired. A rejected candidate can still weigh in on whether your hiring process was worth going through.

The method originates from the Net Promoter Score that Bain & Company introduced in 2003 to measure customer loyalty. Recruiting teams adapted it for two main reasons:

• Leadership already knows how to read an NPS, so the number needs no explanation

• It gives rejected candidates the same voice as the people you hired, who are otherwise the only candidates most companies hear from

Every response falls into one of three categories:

• Promoters (9 or 10): They would recommend that a friend or colleague apply

• Passives (7 or 8): Their experience was fine, but they’re unlikely to go out of their way to recommend

• Detractors (0 to 6): They’re the candidates most likely to advise against applying

A single company-wide score counts an early rejection and a new hire the same. Plenty of teams report that overall number, and I did too, but it gets skewed from both ends: candidates screened out early pull it down, and hires, happy to have the job, push it up. On its own, it tells a recruiter very little about where the experience is going well or poorly.

Most applicant tracking systems can send an NPS survey, so the survey itself isn’t the differentiator. What matters is what you compare the score against. When candidate NPS lives in the same system as hiring data, you can break it down by stage, hiring manager, department, recruiter, or source, and see how it relates to other metrics like quality of hire.

How to calculate candidate NPS

Candidate NPS = % of promoters - % of detractors
Take 200 survey responses: 110 are promoters, 50 are passives, and 40 are detractors. Promoters make up 55% of responses and detractors make up 20%, so your candidate NPS is 55 - 20 = +35.

Passives don’t appear in the formula, but they still count in the total number of responses, pulling both percentages down. A team with a high volume of 7s and 8s can end up with a modest score even when very few candidates are actually unhappy. NPS is not an average rating. If every candidate scores 8, the average rating is 8 and the candidate NPS is 0, because every answer is a passive.

What is a good candidate NPS score?

Any cNPS score above zero means that more candidates would recommend applying than would warn people away. Company and product NPS programs often treat scores of 50 or more as an excellent metric. Candidate NPS usually runs lower, since most respondents were turned down.

The more useful comparison takes into account the furthest stage a candidate reached. A score of 30 from candidates rejected after a first round is a good result. The same score from hired candidates is a warning sign, since hires usually score far higher. It can point to a gap between what the hiring process described and the actual job, or to problems during onboarding. Because hires score so high, report them separately rather than folding them into your main score. Set a baseline from your own data for each stage, then aim to improve that number quarter over quarter.

Kelsey Peterson, VP of Customer Success at Ashby, made the same point at Ashby One 2025: benchmarks are reference points, “less about matching or meeting what we find in trends and more about evaluating where you stand and why.”

Ashby’s candidate NPS benchmarks put the average across customers at 37. Scores ranged from 18 for candidates who left before the onsite to 84 for hired candidates.

Why your score is determined by who answers

Candidate NPS scores are heavily impacted by who is responding. In Ashby’s data, 43% of hired candidates responded to surveys, compared with 27% who reached the offer stage, 21% who reached the onsite, and 16% who left before the onsite.

Because scores differ meaningfully by stage, your overall score depends on the mix of people who answer. It’s a weighted average across stages, so it’s susceptible to how the mix changes, even if nobody’s experience changed. Say early-stage candidates score 15, late-stage candidates score 30, and hired candidates score 85. If responses split 50% early, 30% late, and 20% hired, the overall score is about 34. Now say hiring slows for a quarter and the split shifts to 65% early, 25% late, and 10% hired. The overall score falls to about 26, even though the score at each stage remained the same.

I usually tell people two ways to avoid misreading NPS. One is to report candidate NPS by the furthest stage each candidate reached. Another is to report two scores, one for candidates who reached the onsite or later and one for those who left earlier, which adjusts for changes in hiring patterns quarter to quarter.

Response rates matter as much as the score. The average answer rate across Ashby customers is about 18%, so the vast majority of candidates never answer. Candidate NPS is useful for spotting where to look, but it shouldn’t be the only metric behind a decision. Track response rate next to the score, and look closely at any department where it falls well below your average.

The opposite bias matters too. Unhappy candidates sometimes don’t bother replying. One way to counteract this is by telling candidates in the survey invitation that the recruiting team reads every response and uses it to improve the process, so they know their answer will be read.

How to set up a candidate NPS survey

Ashby’s guide to candidate experience survey design provides an overview of question banks and survey design in depth. For a candidate NPS program, below are a few key decisions to make:

Who to survey: Send the survey to hired, rejected, and withdrawn candidates when they finish the process, and report hires as their own group. Folding them into the main score inflates it, and leaving out rejected candidates skews it the other way.

When to send: Send it 48 to 72 hours after the decision, and never send it before a rejected candidate has heard the outcome. Notion’s recruiting team uses a three-day delay so recruiters can send a personal rejection first. Ashby’s survey data found that 94% of responses arrived within nine days of the request, so close each reporting period about 10 days after the last invitation goes out.

How many surveys: One, sent when each candidate leaves the process. You can filter the results by the furthest stage each candidate reached, so you don’t need a separate survey for each stage.

How many questions: Three at most. Ask the NPS question, plus an optional interviewer rating and comment box. In my 15 years in talent acquisition, the surveys that get answered are the short ones, especially when the candidate has been rejected. Keep the NPS question the same every time so the trend stays comparable.

Anonymous or named: Let candidates choose. Optional anonymity lets candidates who want to stay unnamed answer honestly, and you can still follow up with anyone who puts their name to a bad experience. Ashby’s data shows named-only surveys get slightly more responses (20.6% against 17.7% when anonymity is optional and 15.3% when every response is anonymous).

How candidate NPS surveys work in Ashby

Candidate Experience Surveys are available on every Ashby plan. Organization Admins create and manage them.

Each survey can include one NPS rating question on the standard 0 to 10 scale. Ashby automatically records each answer as a promoter, passive, or detractor, so you don’t have to categorize responses by hand. Each survey also has one of three anonymity settings: no anonymous answers, anonymity as the candidate’s choice, or anonymity for every response.

Ashby’s automation rules send a survey when a candidate meets conditions your team sets, such as being hired, archived, or withdrawing from the process. You can add a delay to the survey email, which is how teams make sure a rejection reaches a candidate before a survey. Surveys can also be requested manually from a candidate’s profile.

The prebuilt Candidate NPS Score report shows your score over time and can be filtered by team. On the Plus and Enterprise plans, custom reports can break the score down by job, department, recruiter, stage, or outcome, in the same system as your hiring data.

Plus and Enterprise customers can also set up detractor alerts: custom reports filtered to detractors, with an alert that notifies the right person when a new low score comes in. I filter mine to detractors who left a comment, since the comment is where the reason shows up.

What to look for in an ATS that measures candidate NPS

Most applicant tracking systems can send a survey. What differentiates them is what they’re able to do with the responses. Here are five capabilities to check for:

1. NPS built into candidate surveys, with automatically calculated promoters, passives, and detractors.

2. Surveys sent automatically when a candidate’s process ends, whether they’re hired, rejected, or withdraw, with a delay you control so the decision reaches them first.

3. The ability to determine if the survey is anonymous, optional, or named responses.

4. Reporting that records the stage each candidate reached when the survey went out, and connects their answers to applicant data such as hiring manager, recruiter, source, and time in process.

5. Alerts when a low score comes in, so it can be quickly analyzed.

Of the five, having unified reporting makes the biggest difference. When surveys live in a separate platform, scores arrive on their own with no application data attached. With both combined in one system, a team can see the stage each response came from and compare scores across hiring managers, recruiters, and sources to find out who and what is working well. It can also set candidate NPS against other metrics, like quality of hire.

How to read your candidate NPS

Most teams start by looking at the monthly trend and dig in further when they see movement. Gorgias and M-KOPA both work this way: when a score shifts, they filter by department, outcome, and recruiter to diagnose the change. Furthest stage and source can be useful cuts as well.

Filtering the scores makes trends and issues easier to spot. Ten complaints about slow feedback may mean little scattered through a quarter’s worth of responses. If all 10 come from engineering candidates rejected after the final interview, the team knows exactly where to look.

The stages I watch most closely are the second round, the final round, and the offer, because that’s where most candidates hear they aren’t moving forward. When candidates turned down at the offer stage still rate the process highly, the process is doing its job.

Separating feedback about the outcome from feedback about the process also helps. Some detractors are unhappy because they didn’t get the job. Others dealt with an unprepared interviewer, didn’t know what the next stage involved, or waited weeks to hear back. That second group is the feedback a team can act on to improve the hiring process.

With an alert on detractors who left a comment, the recruiting team gets a running list of unhappy candidates who explained why. Group those comments by department, stage, hiring manager, and recruiter each month, and send each owner the summary for their area.

How recruiting teams put candidate NPS into practice

Collecting candidate NPS scores shows where candidates are having a bad experience, but it doesn’t tell you what to do about it. A small recruiting team can personalize almost every interaction. A team hiring hundreds of people a quarter likely prioritizes clear ownership and a repeatable way to follow up.

Each of the teams below acts on what candidates tell them in different ways. Some teams make one person accountable for the number, some follow up personally when a candidate has a bad experience, and others focus on the small details of the process that candidates remember.

At Gorgias, one recruiter handles candidate NPS, spending about 20% of their time on it alongside the rest of their recruiting work. They choose the survey questions and have tested sending the survey at different points in the hiring process.

At M-KOPA, the team keeps survey responses named so it can follow up. When a candidate reports that something went seriously wrong, the team can find out what happened and reach out to that person directly.

At Notion, recruiters work from one of the company’s operating values, being kind and direct, and write personal rejection notes rather than relying on automated emails. The team sends its survey after the onsite, and Cuyler Hirata, Head of Recruiting Operations & Analytics, shares the responses with recruiting leadership every month.

At Sierra, the details are built into the workflow. Recruiters use a custom field in Ashby to record each candidate’s favorite snack, which is waiting for them when they arrive for an onsite. The team keeps a channel for shout-outs about good candidate moments and tracks its candidate NPS over time. Sierra’s Head of Recruiting, Greg Marsh, described the approach on Ashby’s Offer Accepted podcast.

Frequently asked questions

How should you benchmark your candidate NPS?
Any score above zero means more candidates would recommend applying than would advise against it. The more useful benchmark is your own score for each furthest stage reached: compare candidates who left before the onsite with the same group last quarter, not with your hires. Report hires as their own group, since they score far higher and push up any combined number.

Can candidate NPS be negative?
Yes. A negative score means detractors outnumber promoters, and it usually points to a process problem, such as a hiring team that isn’t aligned on what it’s looking for, or no clear process for candidates to move through. Candidates turned down at application review or the first screen can score lower, but a negative score there shouldn’t be accepted as normal. A well-run recruiter screen can earn a high score, and recruiters should aim for one.

Should candidate surveys be anonymous?
Let candidates choose. Optional anonymity gets honest answers from people who would rather stay unnamed, and named answers still let you follow up with a candidate who had a bad experience. Fully anonymous surveys get the fewest responses: 15.3% in Ashby’s data, against 20.6% for surveys that require a name.

Which applicant tracking systems measure candidate NPS?
Several applicant tracking systems include an NPS question in candidate surveys. Ashby builds candidate NPS into its Candidate Experience Surveys on every plan, with automated sending by stage or outcome, three anonymity settings, and a prebuilt Candidate NPS Score report all in one system.

Already using Ashby? The Candidate Experience Surveys course in Ashby Academy walks through configuring the survey, automating requests, and reporting on candidate NPS. If you’re evaluating an ATS, request a demo to see candidate NPS reporting alongside the rest of your hiring data.