How to tell an internal candidate they did not get the job

Tell an internal candidate they did not get the job in person or by call, the same day the decision is final and before anyone else on the team hears anything, with a specific, honest reason tied to this role rather than a verdict on their overall performance, and a concrete next step for their development discussed in the same conversation. The reason this conversation is harder than an external rejection is that you still have to work with this person tomorrow, and how you handle the next ten minutes has more influence on whether they stay than almost anything else in the process.

Why this conversation is not a rejection email with extra steps

An external candidate absorbs a no and moves on with your company. An internal candidate absorbs a no and comes back to their desk, on your team or a team you work with, the same afternoon. They will see the person who got the role in meetings. They may train them. Internal rejections carry a second signal external ones don't: a judgment about where this person stands relative to a peer they know, in an organization they've already invested years in.

That's why the mechanics matter more here than the wording. Timing, who delivers the news, and whether the reason holds up against what the candidate already knows about their own record decide whether this becomes a manageable disappointment or the start of their exit.

Before the conversation: five decisions to make first

1. Decide exactly what you'll say was the deciding factor, and check it's true. Go back to the scorecard. If you can't point to a specific competency or piece of evidence, you are not ready to have this conversation yet, because the candidate will ask and a vague answer is what does the damage.
2. Decide who delivers the news. Usually the hiring manager for the role, not the candidate's current manager relaying secondhand, and not HR alone. If the candidate's current manager is a different person from the hiring manager, loop them in beforehand so they're not caught off guard and can support afterward.
3. Decide the order of conversations. Tell the unsuccessful internal candidate first, ideally before you extend the offer becomes public knowledge, and always before a team-wide announcement. Getting this order wrong is the single most common mistake in this process.
4. Decide what you can offer next. A specific development step, a different opportunity, a named gap to close before the next opening. Walking in with nothing concrete to offer turns the conversation into pure bad news with no path forward.
5. Block real time and a private room or call. Fifteen minutes minimum, in person or on video with cameras on if remote, never by chat or email as the first notice. A calendar invite titled vaguely ("quick chat") the same morning is fine; a same-day surprise with no warning at all reads as ambush.

The conversation script

Adapt the brackets. The shape matters more than the exact words: news first, specific reason, no false hope, then forward.

Open (10 seconds, no preamble):
"[Name], thanks for making time. I want to tell you directly: we've
decided to move forward with someone else for [role]. I know that's
disappointing, and I wanted you to hear it from me first."

[Pause. Let them react before you continue.]

The reason (specific, tied to the role, not a verdict on them):
"The deciding factor was [specific competency/evidence]. [Other
candidate's name, once decided] has [specific strength relevant to
this role]. This isn't a reflection of your work on [current team] --
it's specific to what this particular role needed right now."

What you're NOT going to do:
- Don't say "it was really close" unless it's true and you can back
  it up -- false closeness is a common, well-meant lie that backfires
  when the candidate compares notes with the interview panel later.
- Don't list every reason at once. One or two clear, defensible
  points land better than five that feel like a pile-on.
- Don't promise "next time for sure" -- you can't guarantee that and
  it undermines the next opening's real competition.

The path forward (say something concrete, even if small):
"Here's what I think would close the gap for a role like this next
time: [specific, actionable]. I'd like to set up a development
conversation in the next two weeks to build that in -- can we put
that on the calendar now?"

Close:
"I know this stings. Take the time you need to process it, and let's
talk again once that's settled -- I'm not expecting you to have
questions right now, but I want you to know I'm around when you do."

Send a short follow-up within the hour, so the candidate has something to reread once the conversation has sunk in:

Subject: Following up on our conversation

[Name],

Wanted to follow up in writing on what we talked about. I know it's
a tough one, and I meant what I said: [one-line restatement of the
specific reason]. That's specific to this role, not a read on your
work on [team].

I'd like to get [development conversation] on the calendar in the
next two weeks -- I'll send a time, or let me know what works.

I'm around if you want to talk more, today or any time.

[Your name]

Handling the reactions you'll actually get

The reason you give has to survive them comparing notes

Internal candidates talk to their interview panel, to peers, sometimes to the person who got the role. A reason that doesn't match what interviewers actually scored gets discovered, and it costs you more trust than a harder truth delivered honestly the first time. Before the conversation, check your stated reason against the actual scorecard from the loop; see internal interview feedback examples for wording that stays specific and defensible under exactly this kind of scrutiny.

Avoid two traps in particular. The first is vague comfort language: "it was just a tough field" or "everyone was great" tells the candidate nothing they can act on and reads as evasive the moment they think about it. The second is over-explaining: five reasons stacked together feel like a case being built against them rather than one honest factor. Pick the one or two that actually decided it.

A 90-day retention plan after the conversation

The conversation is not the end of the risk. Internal candidates who don't get the role are a measurable flight risk in the months after, and a specific follow-up plan is what determines whether that risk becomes a resignation.

• Week 1: the development conversation you promised, with a written, specific plan: skills to build, a stretch project, or a timeline for the next opening if one is realistic.
• Week 2–4: a check-in that isn't only about the rejection. Normal 1:1 work, so the relationship doesn't become defined by this one event.
• Day 30: a direct check: "how are you feeling about things, honestly." Some people need a month to be ready to answer that truthfully.
• Day 60–90: visible progress on the development plan, even if small. A plan that was promised and then forgotten is worse than no plan at all.

If the candidate reports to a different manager than the hiring manager, make sure that manager owns this plan, not the hiring manager who is now managing someone else. A retention risk that falls between two managers' responsibilities is how good people leave quietly.

Mistakes that turn disappointment into a resignation

A checklist before you have this conversation

• You have a specific, evidence-based reason that matches the scorecard, not a general impression.
• You've decided who delivers the news and confirmed it's not left to a secondhand relay.
• The order is set: this candidate hears it before any wider announcement.
• You have at least one concrete next step to offer, even if the role search is still open.
• You've blocked real time, in private, not a hallway conversation or a chat message.
• A written follow-up is drafted, ready to send within the hour.
• A 30/60/90-day follow-up is on your own calendar, not just promised in the room.

If you run an internal loop with the same rigor as an external one, the reason you give in this conversation should already exist on paper. Interview Signal builds a scorecard with the candidate's own words checked against the transcript during the loop, so when this conversation happens, you're reading from evidence rather than reconstructing a justification from memory under pressure.