Most people start a job search with a tracker and abandon it somewhere in week three. Not because tracking is hard, but because the thing they built asks for more than the search gives back. Twenty columns get filled in carefully for the first six applications and left blank for the next forty, and once a tracker is half-empty it stops being trustworthy, and once it stops being trustworthy there is no reason to open it.
So the useful question isn't what to record. It's what you will still be recording in month two, when you are tired and applying to four roles on a Sunday evening.
Record five things, and only five
Everything below survives a bad week. Anything beyond it is optional and should be treated as optional, which means leaving it blank has to be fine.
- Company and role. Both, always. "Senior Designer" on its own is useless three weeks later when you have applied to two companies with that title, and one of them had two openings.
- The link to the posting. The single highest-value field, and the one most likely to be skipped because it feels redundant at the time. Postings get taken down. When a recruiter calls about a role you applied for five weeks ago, the description is the only thing that tells you what they actually asked for. Save the URL, and if the role matters, paste the description somewhere too.
- The date you applied. Not the date you found it. The gap between those two is what tells you whether silence means anything.
- Status. One word. More on this below.
- A log, not a notes field. The difference matters. See the section after next.
Notice what isn't there: salary, location, the recruiter's name, how you found it, a rating out of five, the cover letter version you sent. Every one of those is useful in the specific moment you need it and dead weight the rest of the time. Add them if you find yourself wanting them twice. Don't add them in advance.
Pick a status set that matches how applications actually end
Most trackers use three or four statuses and then have nowhere to put the majority of applications, which end in nothing at all. That is the failure: a pipeline that models the happy path and leaves the normal path homeless.
A set that covers the real distribution:
- Saved — found it, haven't applied. This column is a to-do list, and it is allowed to be long.
- Applied — sent, waiting.
- Interviewing — anything past a screen, including the recruiter call. It is one column, not five, because a five-stage interview pipeline is a hiring manager's problem, not yours.
- Offer — the one you are doing this for.
- Rejected — they said no.
- Bad fit — you said no, or would. Worth separating from rejection: the ratio between the two is genuinely informative, and collapsing them makes a search look worse than it is.
- Stale — nobody said anything and it has been long enough that nobody will. Most applications end here. Giving that outcome a name is the single change that makes a tracker feel honest instead of accusatory.
Without a stale column, dead applications sit in "Applied" forever and the column becomes a wall of things you are supposedly waiting on. Then you stop looking at it.
Log events, don't maintain a notes field
A notes field is a single blob you overwrite. A log is a list of timestamped entries you append to. The difference shows up about six weeks in, when someone asks "when did you last hear from them?" and a notes field can only answer with whatever you happened to type most recently.
Entries worth appending, each one line:
- Applied — via the company site / a referral from X / a recruiter.
- Recruiter screen, 25 mins. She said the panel is three people and they want to close by the end of the month.
- Sent the follow-up.
- Take-home sent, due Friday.
- Nothing since the 3rd.
That last kind of entry — recording that nothing happened — is what makes the log worth keeping. It is the difference between "I think they went quiet" and knowing the exact date they did.
The other reason to log: interviews reward specificity, and specificity decays fast. "They mentioned they're rebuilding the onboarding flow this quarter" is a strong thing to bring up in a second conversation and an impossible thing to recall unprompted three weeks later.
Capture at the moment of applying, never afterwards
This is the mechanical part, and it is where most systems fail. A tracker you update in a batch on Sunday is a tracker that is wrong all week and eventually wrong permanently, because the batch gets skipped once and then the backlog is too big to face.
The rule that works: nothing counts as applied until it is in the tracker. Make the entry part of hitting submit, not a chore that follows it. That means the capture step has to be fast enough to not be worth avoiding — which in practice means it has to be roughly one action, on the page you are already on, rather than a tab switch, a scroll, a new row and five fields of typing.
The four ways a tracker dies
- Too many columns. Fixed by deleting them. A column that has been blank for twenty rows is not a column, it is a reproach.
- No outcome for silence. Fixed by a stale status and a rule for when things move into it.
- Batched updates. Fixed by capturing at submit time.
- It lives somewhere you don't go. A file on a desktop you only use in the evening loses to a browser tab. Whatever you use has to be somewhere you already are when you are applying.
Reviewing it, which is the actual point
Once a week, twenty minutes. Three questions:
- What needs a follow-up? Anything in Applied past your threshold — see how to follow up on a job application.
- What should move to Stale? Be ruthless. An application you have written off is easier to let go of than one you are vaguely still hoping about.
- Where are the ones going well coming from? If every interview came from a referral and none from a job board, that is an answer about where to spend next week, and it is the only thing in the tracker that changes what you do rather than just recording it.
That third question is the whole return on the exercise. Everything else is bookkeeping in service of it.
If you want to start from a spreadsheet, the columns and formulas worth having are here, along with a straight account of where a spreadsheet runs out.