Gains:
- Ability to produce sourcing, invitation, offer and rejection messages in a professional and respectful manner
- Ability to design timely and personalized communication flows that improve candidate experience
- Ability to interpret recruitment funnel metrics and improve communication accordingly
A candidate accumulates an impression with every contact he or she has with your company during the hiring process: response to the ad, first email, invitation to interview, offer or rejection. The sum of these impressions is called candidate experience and directly affects your employer brand, the number of future applications, and even customer loyalty. A candidate who receives a bad rejection message or one that goes unanswered for weeks describes their experience on and around social media. Artificial intelligence (AI) enables you to produce fast, consistent, and respectful copy at every step of recruiting communication — but you ensure the tone, empathy, and accuracy. In this unit, we will learn to establish the entire communication flow with AI, from sourcing to offer and rejection messages, and to interpret funnel metrics.
Terms: Sourcing is proactively reaching out to candidates who have not applied to the job posting but may be suitable (e.g. LinkedIn message). The recruitment funnel is the stages that candidates go through from application to recruitment: application → pre-screening → interview → offer → recruitment. Conversion rate is the percentage of candidates moving from one stage to the next. Personalization means tailoring the message to the candidate's actual background and role — not copy-pasting.
Map of Communication Flow
A good candidate experience is no accident; At each point of contact, what to say is planned in advance. Step by step:
- List the touchpoints. Sourcing message, application thanks, interview invitation, post-interview information, offer, rejection.
- Set tone and timing for each point. When, in what tone, with what information.
- Generate templates with AI. Keeping each message in brand tone and leaving areas for personalization.
- Personalize and verify. Verify with reality the information that AI can make up, such as name, date, position, etc.
- Measure response time. Don't leave the candidate in limbo; Tell "when you'll hear from me" in every message.
Your role: a recruiter who cares about the candidate experience. Write a personal sourcing (first contact) message for the following candidate. Channel: LinkedIn. Length: 120 words maximum. Include: why I reached out to him (a concrete detail from his profile), a compelling 1-sentence summary of the role, a low-pressure interview invitation. Tone: respectful, friendly, no sales pressure. Leave personalization areas with [SQUARE BRACES].<candidate>Role history: 4 years B2B field sales, fintech industryWe're looking for: Corporate Sales Manager, Istanbul</candidate>
Tip: In each message, give the candidate “what will happen next and when.” Uncertainty is what hurts the candidate experience the most. A single sentence "We will get back to you within 5 business days at the latest" is much better than silence.
Offer and Rejection Messages: The Most Sensitive Moments
The offer message must create excitement; The rejection message should maintain respect. Rejection is the most critical moment of the candidate experience because most candidates get rejected and this is where they form their lasting impression of your company.
Write a negative result (rejection) message to a candidate. Rules:- Short, respectful, tone thanking the candidate for their time.- DO NOT attribute the reason for the rejection to a personal characteristic such as the candidate's age/gender/circumstance; creating legal and ethical risk.- Leave the door open if possible (“we would like to stay in touch for future roles”).- Do not fake praise or empty promises. Length: 90-120 words. Tone: warm but professional.
Write a written offer summary email after the oral offer to the selected candidate. Include: position, start date [ ], work model [ ], salary and benefits [ ], next step (signature/interview), contact person. Tone: warm, clear, celebratory. Leave the numbers and dates blank with [ ]; I will fill it with actual values.
Attention: Never use expressions that imply the candidate's age, gender, marital status, health or a group in the rejection reason. Job-related, neutral statements such as "We moved forward with a more experienced candidate for the position" are both accurate and safe. AI can sometimes create risky justifications by over-explaining; Read every rejection text with this eye.
Weak Prompt / Strong Prompt
Weak prompt: Write a rejection email to the candidate.
The result: a cold, general, copy-paste-feeling text that sometimes creates risks by explaining too much.
Strong prompt: [rejection + short and respectful tone + prohibition of linking the reason to personal characteristics + leaving the door open + prohibition of fake praise + length limit]
The result: a respectful, legally secure and consistent message that protects your brand.
Interpreting Funnel Metrics
The numbers show where the candidate experience weakens. AI is a good helper at interpreting your funnel data and suggesting possible causes and improvements — but you validate the suggestions with field knowledge.
Interpret the hiring funnel data below. Calculate the conversion rate for each stage, highlight the 2 weakest stages, and suggest 2 possible causes + 1 testable improvement for each. Make a definitive diagnosis; Present it as a hypothesis, state that it needs to be verified.<data>Application: 400Preliminary screening: 120Interviewee: 90Offer made: 20Offer accepted: 8</data>
Stage
passing
Conversion
Possible signal of weakness
Application → Pre-selection
120/400
30%
The ad may be attracting off-target candidates
Preliminary selection → Interview
90/120
75%
healthy
Interview → Offer
20/90
22%
Interview/expectation mismatch
Offer → Acceptance
8/20
40%
Salary/competitive offer/process slowness
The two most striking points in this table are the low offer acceptance rate and the low transition from interview to offer. AI flags these; You verify salary band, competitor offers and process speed in the field.
Three Mini Cases
Case 1 — The price of silence. A company wasn't sending messages to candidates it rejected. An employee survey and social media scan showed employer brand scores dropping. A respectful rejection template with personalization area was created with AI, and a return policy of 5 business days was introduced for all candidates. After 4 months, the number of career site applications increased by 17%; Complaints of "I did not receive any feedback" almost disappeared.
Case 2 — Offer acceptance rate. One technology company had a bid acceptance rate of 45%. AI interpreted the funnel data and flagged the hypothesis “long time between offer and first contact.” The process was examined: the offer came in writing an average of 6 days after verbal approval. When the deadline was reduced to 2 days and the offer email was rewritten in a warm tone, the acceptance rate increased to 63%.
Case 3 — The power of personalization. A sourcing team was sending copy-paste LinkedIn messages and getting a 6% response. A template was set up with AI that added an actual detail from the candidate's profile ([square brackets] should be filled in) to each message. The team filled in the details by hand; the response rate increased to 14%. The key was to do true personalization without letting the AI make it up.
Common mistakes
- Leaving the candidate in limbo. Not telling "when you'll hear back" is the mistake that hurts the experience the most.
- Attributing the reason for rejection to personal characteristics. Justifications that imply age, gender, or status are both unethical and legal risks.
- Copy-paste sourcing. Mass messages without personalization will receive low response and wear out the brand.
- Sending AI-made up information without verifying it. A message sent with the wrong name, position or date undermines trust.
- Not measuring metrics at all. Without funnel data, you can't see where the experience is weakening.
- Mistaking an AI recommendation for a definitive diagnosis. Suggestions are hypotheses; verified by field and data.
In summary
Candidate experience is the sum of impressions accumulated at every touchpoint and directly determines your employer brand. AI; It produces sourcing, invitation, offer and rejection messages quickly, consistently and respectfully, and interprets funnel metrics. But you ensure the tone, the empathy, the accuracy, and the final decision. The biggest difference comes from personalization and timely communication that eliminates uncertainty.
Application task
Design the communication flow of a position. (1) Generate brand-tone templates with AI for six touchpoints (sourcing, thank you, invite, inform, offer, rejection), leaving the personalization fields [SQUARE BRACKETS]. (2) Produce the rejection template with the prohibition of "attributing the reason to personal characteristics" and read it against legal risk. (3) Have the AI interpret a sample funnel data and extract the two weakest stages and improvement hypotheses. (4) Write how you would verify a hypothesis in the field.
checklist
- [ ] Have the tone and timing been defined for each touchpoint?
- [ ] Does each message provide the candidate with "next step and deadline" information?
- [ ] Is the rejection message respectful and reasoned/neutral?
- [ ] Are the personalization fields filled with real information (not made up)?
- [ ] Have funnel metrics been measured and interpreted?
- [ ] Have the AI's suggestions been taken as hypotheses and validated?