Each week, track five numbers: how many applications you submitted, how many replies you got per 100 submissions, how many of those replies became interviews, why jobs were skipped or paused, and how your results differ by role title. Together they tell you whether to change your resume, your filters, or your volume. Everything else is optional. The aim is a short weekly review that leads to one concrete change, not a dashboard you stare at every day.
Why weekly, not daily
Recruiters do not reply on your schedule. An application sent on Monday might get a reply on Thursday or the following Tuesday, and a single day of data can swing wildly because of one lucky or unlucky posting. A week is long enough to smooth out that noise and short enough that you can still correct course before a month is gone. Pick a fixed day, Sunday evening or Monday morning, and do the review in the same order every time.
Metric 1: Submitted applications
Count only applications that were actually submitted. Skipped and paused jobs are not applications. In the AutoApply CV applications tracker, each job lands as submitted, skipped or failed, so this number comes straight from the submitted count. Write it down even if it is small. A free plan user applying 3 times a day will see around 21 in a week, and that is enough to start spotting patterns if the targeting is tight.
Metric 2: Reply rate per 100 applications
A reply is any human response: a recruiter email, a LinkedIn message, a call, a screening test link, or even a polite rejection written by a person. Automated "we received your application" emails do not count.
Normalize the number to "replies per 100 applications" so weeks of different sizes are comparable. If you sent 40 applications and got 2 replies, that is 5 per 100. If next week you send 25 and get 2, that is 8 per 100, which is a real improvement even though the raw count did not move.
Do not compare your number with figures you read online. Reply rates depend heavily on role, seniority, market and city. Compare yourself with your own previous weeks.
Metric 3: Reply-to-interview conversion
Of the replies, how many turned into an actual interview round? If you get replies but few interviews, the problem is usually after the application: the screening call, your stated salary or notice period, or a mismatch the recruiter noticed when they spoke to you. If you get almost no replies at all, the problem is upstream, in the resume or the jobs you are choosing.
Metric 4: Skip and pause reasons
This is the metric most people ignore, and it is the cheapest to act on. Group your skipped and paused jobs by reason:
- External apply: the job sends you to a company site. If this is a large share of your search, your filters may be pulling in roles that are not Easy Apply. These are worth a manual application if the role is strong.
- Already applied: duplicate prevention caught it. A high count means your searches overlap heavily, and you can broaden titles or locations.
- Validation pause: LinkedIn rejected an answer, such as a decimal in a whole-number field. Each repeated pause points to one answer you should fix in the answer bank.
- Unknown question: a screening question the answer bank could not answer truthfully. Add an answer once and it stops costing you jobs.
None of these are charged against your allowance, but each one is a job that did not get an application. Fixing the top reason each week usually gives you more submissions without changing anything else.
Metric 5: Results by role title
Split submitted applications and replies by the title you targeted. You might find that "Backend Engineer" gets replies while "Full Stack Developer" gets none with the same resume. That is a signal either to drop the weaker title or to make a version of your resume that genuinely fits it. Our guide on choosing skill tags that convert covers how to adjust keywords per title.
A simple weekly log
You do not need a spreadsheet template. A note with these lines works:
- Week of: date
- Submitted: number
- Replies: number, and replies per 100
- Interviews: number
- Top skip or pause reason: one line
- Best and worst role title: one line each
- One change for next week: one line
The last line is the point of the whole exercise. If you cannot name one change, you have not finished the review.
How to read the numbers together
- Low submissions, decent reply rate: your targeting works. Increase volume, either by widening locations or by moving beyond the free daily limit. See pricing for options.
- High submissions, very low reply rate: volume is not the issue. Tighten filters and rework the resume before sending more.
- Many pauses: spend ten minutes on the answer bank before the next run.
- Replies but no interviews: review what you are saying on screening calls and in salary or notice period fields.
Change one thing at a time
If you rewrite your resume, change titles and switch locations in the same week, you will not know which change helped. Make one change, run a full week, and compare. It feels slower, but it is the only way the tracking actually teaches you something.
Next step
Open your applications tracker now and fill in last week's seven lines, even if the numbers are small. Then pick the single most common skip or pause reason and fix it before your next run.
