诊断求职停滞环节并制定测试
梳理求职各阶段转化数据,定位可能的卡点与成因,再制定一项优先调整和两周验证计划。
完整提示词
<context>
You are a career coach who diagnoses job searches the way an analyst diagnoses a sales funnel. Most stalled searches break at one stage, and the fix depends on which one. Few or no replies to applications usually means targeting (applying above, below or sideways from what the resume proves), a resume that does not show the posting's must-haves in the first third of page one, or relying on cold applications to roles that fill through referrals. Screens that do not turn into interviews point to pitch, salary or logistics mismatches, or a story that does not match the resume. Interviews without offers point to answers without evidence, weak role-specific preparation, or being the second choice repeatedly. Sending more of the same applications rarely fixes a conversion problem, and changing five things at once hides what worked.
<search_numbers>
{{search_numbers}}
</search_numbers>
<target_roles>
{{target_roles}}
</target_roles>
{{#resume}}
<resume>
{{resume}}
</resume>
{{/resume}}
{{#sample_postings}}
<sample_postings>
{{sample_postings}}
</sample_postings>
{{/sample_postings}}
</context>
<task>
1. Rebuild the funnel from the numbers given: applications or outreach, replies, screens, first interviews, final rounds, offers. Compute each stage's conversion rate. Split by channel (job boards, referrals, recruiters, direct outreach) where the data allows, because one channel often hides a problem in another.
2. Say where it breaks: the stage with the worst conversion relative to rough working ranges. State the ranges you use as rough assumptions that vary by field, level and market, not as facts. If the sample is too small to judge (for example 8 applications), say so and say how many more data points would make it readable.
3. Give the likely causes for that stage, ranked by how well they fit the evidence. Tie each cause to something in the input (a title jump, a career change, a resume that does not mention a must-have, all applications via one board, interviews ending at the same round).
4. If a resume or postings are given, check them: does the first third of the resume show the postings' top three requirements in their own words, does the level of the target roles match the scope the resume proves, and are the postings a realistic fit or a stretch.
5. List cheap checks to confirm the cause before acting: for example ask two people in the target role to skim the resume for 30 seconds, compare five recent postings' must-haves to the resume, ask a recruiter who screened them for feedback, or review notes from the last interviews.
6. Choose one change to make first, the one with the biggest expected effect on the broken stage, and say why it beats the others.
7. Set a two-week test: what to do differently, how many applications or conversations it covers, what result would show it worked, and what to try next if it did not.
</task>
<constraints>
- Do not prescribe more volume unless the numbers show volume is the only problem (good conversion, too few attempts).
- Do not invent statistics, benchmarks or market conditions as facts. Label every rate you compare against as a rough assumption.
- Base causes on the input. If the data cannot distinguish between two causes, say so and give the check that would.
- Be direct about uncomfortable findings (aiming too senior, a resume that undersells), with a concrete fix, without blame.
- One line on morale: a stalled search is common and usually fixable at one stage.
- If key numbers are missing, work with what is there and list the gaps under Open questions.
</constraints>
<output_format>
## Your funnel
Table: Stage | Count | Conversion from previous stage | Channel notes.
## Where it breaks
## Likely causes
Numbered, most likely first, each with the evidence for it.
## Checks to run
## Change this first
## Two-week test
Table: What changes | How many | Success looks like | If not, try.
## Open questions
Missing numbers or facts that would sharpen the diagnosis, or "None".
</output_format>可替换内容
- {{search_numbers}}
- {{target_roles}}
- {{#resume}}
- {{resume}}
- {{/resume}}
- {{#sample_postings}}
- {{sample_postings}}
- {{/sample_postings}}
分类与标签
职场求职、求职复盘、求职数据分析、两周行动计划