A Lightweight Tracker for Job-Search Decisions in China

Record the facts that change your next move, not the number of applications

Most people start a job search in China by counting: how many applications went out this week, how many are still open, how many interviews came back. The number feels like progress and answers nothing. It tells you that you have been busy; it does not tell you whether what you sent was the right thing, whether the employer is still deciding, or whether you are waiting on a process that already ended.

A tracker worth opening holds a different kind of entry. Not "I applied to a dozen places." Instead: which version of your materials went out to this role, what the employer or the platform told you to expect next, the date the silence began, and what you decided the last time you sat in this exact position. Those entries turn a Tuesday evening into a decision instead of a worry.

The test for any line you add is simple: would knowing this change what I do next? If yes, write it. If not, it is diary material. A polite "we will keep your application on file" that comes with no date, no request and no next step changes nothing, so it does not earn a row. The file serves the version of you who has to act tomorrow, not the version who wants proof tonight that effort was made.

The thing you are shopping for is alignment evidence: which document answered which requirement in which posting. Once that line exists, a stalled round becomes answerable — was the problem the fit, the timing, or the document itself? Without it, every quiet week reads identically and teaches you nothing.

A municipal page aimed at students in Shanghai makes the same point about materials, from the other direction. How detailed should your resume be? describes a recent master's graduate, Li, who was told a well-crafted resume was the key to landing a job in China. She compiled her academic and internship experiences, used AI tools to polish the wording and picked an attractive template, producing a full packet: cover page, cover letter, three-page resume and back cover. Employers showed little interest and she rarely got interviews even for roles that matched her qualifications. The page's own read is that the dense, unfocused three-page document made it hard for recruiters to see her fit quickly, and its advice is that experience should highlight the duties and skills that align closely with the position, with the superfluous detail cut.

Hold that example at its actual size. It is one municipal portal page written for students and job-seekers, it is not a market rule, and it does not represent what employers generally require or how they generally behave. Treat it as one source's illustration of over-packing, nothing broader. What transfers to your tracker is the underlying habit: record which version of your materials responded to which requirement, rather than how much material you produced. Every time you rewrite something for a posting, leave one line saying what changed and which requirement it was answering. In a month you will not have to guess which version worked — the file will tell you.

The five fields, and how they behave in a real search

Five fields are the floor for every round: the role, the date you sent materials, the version of those materials, what the process told you to expect, and your next action date. Fewer than five and you lose the thread; more and you stop updating it.

• Role. Enough to tell two similar postings apart if you are tracking several at once.
• Date sent. The anchor that turns "a while ago" into "eleven days, and nothing."
• Materials version. A short label, e.g. "B — two-page industry CV, no cover page, tailored summary". This is the field most often skipped and the most useful later.
• What the process told you. Not every email — only what defines your waiting and your preparation: a screening call was mentioned, a written test comes first, an extra certificate was requested.
• Next action date. The day you will do something if nothing arrives. This is the only exit from the row. A record with no next action is an archive, not a plan.

Here is one continuous sample, not a pile of disconnected cells. Take a candidate who sends an application to a Shanghai-based role through a local platform in October 2026 and also has a headhunter circulating the same position. The five fields, filled honestly over about five weeks, read as follows.

• Role: International marketing associate, Shanghai-based consumer brand (platform posting #4471).
• Sent: 2026-10-08.
• Materials: version B — two-page English CV with a one-paragraph tailored summary, no cover page (version A was the full four-page academic CV).
• What the process told me: platform auto-reply mentioned an HR screening call; a headhunter separately said the hiring manager wants a short English CV in PDF.
• Next action date: 2026-10-16.
• 2026-10-09: headhunter confirms the manager read version B. No interview date yet. Next action date unchanged.
• 2026-10-17: action date passed, no screening call. Decision: send one short follow-up via the headhunter, note the date, and set a stop line — if nothing by 2026-10-27, this round is closed as dormant, not chased further.
• 2026-10-24: platform sends a template rejection — "position filled". Same row updated: outcome closed, reason given by template, no interview. Round ends.

That single row now answers three questions a count never could. Which version was in front of the manager? B, the tailored two-pager. Which channel spoke to whom? The headhunter carried the manager's view; the platform carried the rejection. And was the issue fit or materials? The manager engaged with B and then the role closed, so the evidence points away from the document. Nothing in that entry required more than a minute to write.

What you should not record: every message, every tweak, every internal doubt. Anything you will never reopen does not belong. A tracker you do not reopen is a diary nobody reads, and a diary nobody reads is a chore that produces no decision.

A rejection is one more line in the same file

Put rejections in the same file as the applications, and write the reason whenever one is given — even a template reason. A rejection you can date and attribute is data. A rejection that lives only in your memory is a mood, and a mood cannot be compared across months. Keeping a separate "failures" note because rejections sting guarantees you will open only the wins and stop learning.

Sort rejections by source, because the source changes what the line is worth. A template email usually signals batch processing and tells you the pool closed. A recruiter's direct message may carry one specific sentence you can act on. A headhunter's account is second-hand and should be marked "per headhunter", because it is the headhunter's paraphrase, not the employer's words. One line each is enough; you do not need an essay. If no reason was given, write "reason unknown" rather than inventing one — a fabricated reason is worse than a blank, because you will act on it.

The same discipline the Shanghai portal page applies to resumes applies to the file. Li's packet — cover page, cover letter, three-page resume, back cover — suggests that more material is not automatically better, and by the same logic a rejection log is not improved by storing every comforting sentence you received. Again, that page is a single page of guidance for students, not a rule about employers; take one demonstration of excess from it, not a generalisation. A rejection is worth exactly one row with a date and a source. That is the version you will still be able to read in three months.

More entries is not a better search: the weekly review

One entry per round, not per message. If a single entry takes more than a minute you are recording too much. The file should be glanceable enough that you actually open it, which rules out daily maintenance, colour-coded status labels and screenshots of every email. That is not job-searching, it is producing a report about job-searching. When you notice that tidying the file is taking longer than tailoring materials or sending applications, that is the signal to cut it back to the five fields.

Once a week, at a fixed time, open the file and make exactly one decision. Pick the row where the next action date has passed or silence has run longest, and write the single change it calls for. One change per week beats a monthly summary nobody acts on, because the effect of that one change is observable by the following review. The decision table below is the whole method, applied to four rows in the same sample search.

The one thing to avoid is turning the review into a rewrite: re-editing old entries, polishing wording, reorganising columns. That feels productive and changes nothing. Also avoid the opposite failure — firing off five follow-ups at once because silence is unbearable. That floods the people you need and blurs which action produced which result. One action, so next week can read the outcome.

There is a state where the review should stop rather than produce a decision: when every open row is either dead by your own choice or blocked on information only the employer or an official channel can give you. In that state the file's job is to hold the open questions and wait, not to manufacture motion. Many searches do not collapse under rejection; they stall because the candidate, with nothing new to act on, no longer knows what to do. Keeping a small file of live questions is how you avoid that particular stall.

What this method cannot do, and what to verify before you rely on it

The tracker manages your own decisions under uncertainty. It is not a substitute for verification and it is not a measure of your worth. It stops being useful the moment you ask it to answer questions that are not yours to settle.

Whether a degree earned outside China is recognised for a particular role, whether a given visa or work-permit path applies to you, what a labour contract may or may not contain, and whether an employer's stated requirement is lawful — none of these can be settled inside your own document. Your file can hold the question and flag it; it cannot answer it. For all of it, verify with the employer or with the relevant official channel, and treat any single page of advice, including the municipal page cited above, as one source rather than a rule.

Before you lean on the five-field method, this is the short list to take to the employer or an official channel — asked once, not at every step:

• For this role, what does the employer actually require at first contact: a short CV, a fuller packet, or something else, and is that stated anywhere you can check?
• Who runs the first screening — the hiring department or an HR function — and do they expect different documents?
• If a round ends without interview, is a reason given, and through which channel does it arrive?
• Will materials in English be accepted, or is a Chinese version expected at some stage?
• Are qualifications earned abroad recognised for this role, and through which process would that be confirmed?
• Which official channel can confirm permit or contract conditions that affect your start date?

Everything else stays in your own file, which should remain short, dated and pointed at your next action.