A recruitment team updates every candidate by tying a message to each stage change in the ATS, letting AI draft a short personal update from the candidate's record, and having a recruiter approve the batch before it goes out. Replies are sorted by AI and handed to the right recruiter, so nobody is left waiting in silence.
#Key takeaways
- Candidates are ghosted because updates depend on a recruiter remembering, not because recruiters do not care.
- Every ATS stage change should trigger a drafted update, including rejections.
- AI writes the personal line from the record; a recruiter approves the batch and owns the tone.
- Replies are sorted into questions, withdrawals and availability changes, then routed to a person.
- For early-stage rejections at volume, your ATS's own templates are usually enough.
#Why do candidates get ghosted when nobody means to?
A recruiter carries thirty or more open candidates across several roles. The client goes quiet after interviews, the recruiter has nothing new to say, and so says nothing. Candidates chase on WhatsApp, LinkedIn and email, and each chase is answered by hand, if at all.
In a Greenhouse survey reported by UNLEASH on the Greenhouse state of job hunting report, december 2024, 61% of job seekers said they had been ghosted after a job interview. For an agency, every one of those candidates is someone it may want to place next quarter, and a talent pool that feels ignored stops answering calls.
#What does a good update look like?
- It says where the candidate stands, plainly: still with the client, not progressing, or moving to the next round.
- It says what happens next and when the candidate will hear again, even when the answer is 'we are still waiting'.
- It has one personal line drawn from the record, such as the role, the interviewer or something they asked.
- It is in the language the candidate used with you, Arabic or English.
- It gives a way to reply that reaches a person.
#How does the update workflow run?
- map stages to messages: for each ATS stage, write the update a good recruiter would send, with the facts it must include.
- trigger on change: when a stage changes, or a candidate has gone quiet past your own threshold, AI drafts the update from the record.
- approve in batches: the recruiter reviews the drafts for their candidates in one sitting, edits and sends.
- sort replies: AI labels incoming replies as a question, a withdrawal, an availability change or something else, and routes each to the owner.
- log: the ATS shows what was sent, who approved it and what came back.
This is a workflow, not an agent: the steps are fixed and the model drafts inside each one. Anthropic, december 2024 recommends starting with the simplest approach that works, and candidate updates rarely need more than that.
#What should a recruiter never hand to AI?
Feedback that could be read as a reason for rejection based on age, gender, nationality, health or family status. Offers and counter-offers. Any message to a candidate who is upset or has complained. The draft can suggest wording, but a person writes and sends those, and the record shows it.
#When are ATS templates enough?
If most of your volume is early rejections for high-applicant roles, the automated emails in Greenhouse, Workable or Bullhorn already do the job. Set them up, make the wording kind and specific, and you have fixed most of the silence without new tools.
Drafted personal updates matter after the first interview, when candidates have invested time and expect to be known. That is also where agencies win or lose the next placement.
#Where does Stennir fit?
Stennir uses AI consulting to check whether candidate updates are the right first workflow against your other admin, such as notes or submissions. AI implementation builds the stage triggers, drafts and reply sorting into your ATS, email and WhatsApp Business, in Arabic and English, with a recruiter approving every batch.
AI training then runs a cohort for recruiters on approving drafts fast without letting the tone slide.
We are
Stennir builds update drafting and reply sorting on your ATS, with recruiters approving every message.
We aren't
Stennir is not a chatbot that talks to candidates on its own, and it does not decide who progresses.
#What should a recruitment lead do next?
Pull every candidate past first interview and count how many have had no update since their last stage change. That number is your baseline. Read AI for operations and the guide to a WhatsApp AI agent in Arabic. Then book a 30-minute call or take the free readiness assessment.
#Common questions
Can updates go out on WhatsApp?
Yes, through the WhatsApp Business platform with approved message templates and the candidate's consent to be contacted there. Many Gulf candidates prefer it to email. The ATS stays the record, so every WhatsApp message and reply is logged against the candidate, not lost in a recruiter's phone.
Will candidates know a message was drafted by AI?
The recruiter approves and sends each one, so it is their message. If you want to disclose AI drafting, add a line to your candidate privacy notice. What matters to candidates is that the update is accurate, kind and answered when they reply, and that a person stands behind it.
What do you need from us to start?
A recruitment lead who owns the outcome, ATS access through its API or a sandbox, your current templates, a list of stages and the facts each update must include, and a sample of anonymised past messages that recruiters think went well. Stennir also asks how you handle candidate consent.
What could go wrong?
A draft can state the wrong stage if the ATS is out of date, or sound cheerful when the news is bad. The defences are stage data checked before drafting, tone rules per message type, recruiter approval of every batch, and a short weekly review of complaints and unanswered replies.
