Candidate Experience Best Practices: The 7 Moments That Decide It (With the 2025 CandE Numbers)

Candidate experience is decided at seven moments: the application form, the acknowledgement, scheduling the first interview, the interview itself, the wait afterwards, the rejection, and the offer. Survale's 2025 CandE Benchmark Research (more than 66,000 candidate surveys) and Cronofy's 2024 survey of 12,000 candidates put a number on each of them, and at every one of the seven the fix has the same shape: a message that goes out when a stage changes, so the candidate hears something without a recruiter having to remember to write it.

What do candidates actually complain about?

Not the tone of the emails. The absence of them.

Survale's 2025 CandE Benchmark Research (vendor data; the 14th edition, collected between 31 January and 31 August 2025 across 110 organisations registered worldwide; the regional split of the sample is not published, and many headline figures are reported for North America only) reports that the top negative comment theme from candidates was lack of communication, ghosting and no follow-up. Candidate resentment reached record global highs: 17% in EMEA, 15% in North America, 12% each in APAC and Latin America, and 2025 was the first year EMEA resentment exceeded North America's (Survale, 2025 CandE key statistics).

Cronofy's Candidate Expectations report 2024 (vendor data; 12,000 candidates in the UK, France, Germany, the USA, Spain, Canada and the Netherlands) asked what the most frustrating part of a process was. The top answer, at 28%, was poor responsiveness or communication (Cronofy, Candidate Expectations 2024). The 2023 Talent Board CandE benchmark, as reported by SHRM, put the same thing first among reasons candidates withdraw: poor communication or no feedback, then a process that takes too long, then salary (SHRM, April 2024).

Three datasets, the same complaint. That is why the table below is organised around silence: where it happens, how long it lasts, and what closes it.

Which 7 moments decide candidate experience?

Every number in the table is vendor data. Many of Survale's figures are reported for North America only; Cronofy's cover seven countries in Europe and North America.

#	Moment	What the data says	The one automation that fixes it	Source
1	Applying	62% of candidates said the application took 15 minutes or less, up from under 40% in 2020. 67% would be annoyed at repeatedly filling in their details to arrange interviews.	A CV upload that fills the profile, a form with a handful of fields, and a confirmation that names a reply date.	Survale 2025, Cronofy 2024
2	The acknowledgement	55% received an automated thank-you after applying. 31% of North American candidates had heard nothing 1-2+ months later. North American CandE winners kept 61% fewer candidates waiting 2+ months.	An acknowledgement sent on application that states who reads it and by when, plus a flag on any application that has sat past that date.	Survale 2025
3	Scheduling the first interview	Only 9% had a first interview scheduled within a day; the largest group (31%) waited 2-3 weeks. 42% had left a process because scheduling took too long. 36% would disengage at the one-month mark.	Self-scheduling slots sent the moment a candidate enters the interview stage, with the interviewer's calendar already checked.	Cronofy 2024
4	The interview	65% of North American employers ran structured interviews; 71% of CandE winners did, and winners scored 13% higher on perceived interview fairness. Resentment rose 70% among candidates given no interview preparation or follow-up.	A role-specific question set delivered to the interviewer and a preparation message delivered to the candidate, both triggered by the booking.	Survale 2025
5	The wait after the interview	60% of the top 10 North American winners dispositioned candidates within 3-5 days of application, 16% more often than all participating employers. 12% of candidates would disengage after one week of waiting, 36% after a month.	A scorecard nudge to the interviewer while the card is open, and a status message to the candidate on every stage change, including "no change yet, decision by Friday".	Survale 2025, Cronofy 2024
6	The rejection	56% received no feedback after being rejected during screening or interviews. 33% were told from a do-not-reply address. Only 23% of North American finalists received job-fit feedback. Specific feedback raised willingness to refer others by 69%.	A rejection drafted from the scorecard, with one concrete reason, sent from a monitored address within a fixed number of days of the decision.	Survale 2025
7	The offer	60% received an offer within one week of the final interview. Offers made within a week raised willingness to refer by 108% and willingness to deepen the relationship by 120%. 95% of candidates accepted the offer presented to them.	An offer workflow that starts when the last scorecard is submitted, not when someone gets round to it.	Survale 2025

How do you fix each of the 7 moments?

1. Applying: cut the form, keep the confirmation

The form has already got shorter: 62% of candidates in Survale's 2025 research said it took 15 minutes or less, up from under 40% in 2020. What has not improved is the duplication. Cronofy found 67% of candidates would be annoyed at repeatedly filling in their details to arrange interviews, and a form that asks for the last three roles after a CV upload triggers the same irritation one step earlier.

The fix is a form with a handful of fields and a CV upload that fills the profile. In Pickr every published job gets a short form (name, email, LinkedIn profile, CV, your consent text), the CV is read into the candidate record on submission, and the applicant gets a confirmation email with a link to check the status of their application. The drop-off points are in why your career page is costing you candidates.

2. The acknowledgement: name a date, then flag anything past it

55% of candidates in the 2025 CandE research got an automated thank-you after applying. Then 31% of North American candidates heard nothing for one to two months or more, while the North American CandE winners kept 61% fewer candidates waiting past two months. The thank-you is not the problem. The thank-you followed by nothing is.

An acknowledgement should say who reads the application and by when. The automation that matters is the second half: a flag on every application that has sat past the date you promised. Pickr's daily brief lists candidates that have been in a stage too long and the stages where applicants drop off, so the promise has a deadline somebody sees.

3. Scheduling: send slots the moment the stage changes

In Cronofy's 2024 report only 9% of candidates had a first interview scheduled within a day, the largest group (31%) waited two to three weeks, and 42% had walked away from a process because scheduling took too long. Candidates do not expect same-day: only 5% expected a day or less, 29% expected about a week and 34% expected two to three weeks. A week is a reasonable target; the data says many processes take two to three.

The automation is self-scheduling: when a candidate enters the interview stage, they get a link to pick a slot from the interviewer's real availability. In Pickr that sits on the candidate's portal page, where they can book interview slots and upload documents, and the calendar invite goes out from the stage change itself. Candidates want this: 57% preferred an automated scheduling system over back-and-forth emails, and 37% named interview scheduling as the top area to automate.

4. The interview: structure for the interviewer, preparation for the candidate

Two numbers in the 2025 CandE research point at the same fix. Structure: 65% of North American employers ran structured interviews, 71% of winners did, and winners scored 13% higher on perceived interview fairness. Google's re:Work guide reports the same direction from its own surveys, where rejected candidates who had a structured interview were 35% happier than those who did not (re:Work, structured interviewing guide; Google's internal data, no methodology published). Preparation: resentment rose 70% among candidates who received no interview preparation or follow-up.

Both are triggered by the booking. The interviewer gets a role-specific question set and the previous stage's notes; the candidate gets a message that says who they will meet, what format to expect and how long it takes. In Pickr the question set is prepared for the role rather than pulled from a generic list, the interview is transcribed live, and the scorecard drafts itself from what was said. One more line from the same data: 27% of candidates were told the salary during the interview without asking, and they rated interview fairness 25% higher. Put the range in the preparation message.

5. The wait after: nudge the scorecard, message the candidate

The silence after an interview is where most ghosting starts; the mechanics are in candidate ghosting: why it happens and how to stop it. The 2025 CandE data puts a benchmark on decision speed: 60% of the top 10 North American winners dispositioned candidates within three to five days of application, 16% more often than all participating employers. The same discipline is what closes the gap after an interview. Cronofy's candidates were patient but not infinitely so: 12% would disengage after a week of waiting and 36% after a month.

The decision is usually not what is late. The scorecard is. The automation has two halves: a nudge to the interviewer while their scorecard is outstanding, and a status message to the candidate on every stage change, including the honest one, "no decision yet, you will hear by Friday". In Pickr the interviewer nudge is drafted when the candidate moves stage, and because the scorecard is pre-filled from the transcript, confirming it takes minutes, so it actually gets submitted.

6. The rejection: one reason, a monitored address, a deadline

56% of rejected candidates in the 2025 CandE research got no feedback at all, 33% were notified from a do-not-reply address, and only 23% of North American finalists received job-fit feedback. The same research shows what feedback buys: specific feedback raised willingness to refer others by 69%, and candidates invited to give feedback after interviews were 106% more likely to refer.

The automation is a rejection drafted from the scorecard, with one concrete reason the interviewer actually recorded, sent from an address a human reads, within a fixed number of days of the decision. In Pickr the follow-up is drafted when a candidate moves stage, in your tone, and nothing goes out without a person approving it. Whether to give a reason at all has legal edges in some jurisdictions, and that part is for your counsel, not an ATS. This is not legal advice.

7. The offer: start the workflow when the last scorecard lands

In the 2025 CandE research 60% of candidates received an offer within one week of the final interview, and offers made within a week raised willingness to refer by 108% and willingness to deepen the relationship with the employer by 120%. 95% of candidates accepted the offer presented to them, which tells you the loss is not at the acceptance. It is in the days before the offer exists.

The automation is a trigger: the offer workflow starts when the last scorecard is submitted, not when a recruiter notices the panel is done.

Does automation make candidate experience worse?

It depends on what you automate. Cronofy's candidates preferred automated scheduling (57%) and wanted more of it, and 81% said a human touchpoint in the process was vitally important. Survale's 2025 research found 39% of North American employers using AI recruiting technology, and complaints about AI or automated screening among the top three negative comment themes.

So: automate the status, keep the judgment human. The acknowledgement, the slot booking, the "no decision yet" message and the offer trigger are status. The interview, the feedback and the offer conversation are judgment. In Pickr the stage-change automations run on rules you set once, the drafted replies wait for a person to approve them, and no hiring decision is automated. The communication feature page shows how that looks in a thread.

What should you measure?

Five numbers, all derivable from stage timestamps:

1. Time from application to the first message a human wrote.
2. Time from interview stage entry to a booked slot.
3. Time from application to the first disposition (advance or reject), with the CandE winners' 3-5 days of application as the bar, and the same clock restarted after every interview.
4. Share of rejections that carry a reason.
5. Time from final interview to offer sent, with one week as the bar.

Stage drop-off sits underneath all five. If candidates leave at the stage with the longest silence, the silence is the cause.

What changes when you are an agency?

The silence is often not yours. The client has the scorecard, the client has not decided, the client is on holiday. The candidate does not know that. From our work at ScalingPPL, the candidate blames the agency, never the client, and remembers the agency's name when the next recruiter calls.

The agency fix is the same automation pointed at the client: a shortlist the client reviews in a portal, with ratings and comments that land on the candidate record, instead of a CV attachment in an inbox. In Pickr that client feedback arrives on the candidate's record, the candidate sees where they stand on their own branded page, and the client update is drafted when the candidate moves stage. The candidate still waits for the client. They just stop waiting in the dark.

How much weight should you put on vendor data?

Some, not all. Survale sells a candidate-feedback platform and has owned the CandE program since 2025, so the organisations in its sample registered to be measured, and many of its headline figures are reported for North America only (Survale, 2025 CandE key statistics). Cronofy sells scheduling infrastructure, interview scheduling included, and its report finds that scheduling matters (Cronofy, Candidate Expectations 2024). Neither makes the numbers wrong, but treat them as direction and measure your own process against the five numbers above.

One more piece of context: Ashby's EMEA Recruiter Productivity report (vendor data from more than 180,000 jobs in EMEA, 2021 to mid-2026) puts the typical EMEA hire at 35 days, with 3.9 interview events per hired candidate by Q1 2026 (Ashby, EMEA Recruiter Productivity). Thirty-five days and four interviews leave plenty of room for silence. That is where the seven moments live.

Sources

• Survale, 2025 Candidate Experience (CandE) Benchmark Research: Key Statistics. https://survale.com/2025-candidate-experience-cande-benchmark-research-key-statistics/ (accessed 2026-10-14; vendor data)
• Cronofy, Candidate Expectations and Preferences report 2024. https://www.cronofy.com/reports/candidate-expectations-report-2024 (accessed 2026-10-14; vendor data)
• Roy Maurer, SHRM, "Research Reveals Candidates' Frustrations with Hiring Process", reporting the 2023 CandE Benchmark Research, 9 April 2024. https://www.shrm.org/in/topics-tools/news/talent-acquisition/candidate-experience-talent-board-research-candes (accessed 2026-10-14; secondary report of a gated primary)
• Google re:Work, A guide to structured interviewing for better hiring practices. https://rework.withgoogle.com/intl/en/guides/a-guide-to-structured-interviewing-for-better-hiring-practices (accessed 2026-10-14; Google internal data)
• Ashby, EMEA Recruiter Productivity, Talent Trends Report, 9 September 2026. https://www.ashbyhq.com/talent-trends-report/reports/emea-recruiter-productivity (accessed 2026-10-14; vendor data)