Why your job search is not working: find the stage that fails, then fix that stage

What is the job search funnel?

The job search funnel is the chain of conversions from the applications you send to the offers you receive. Stage 1 is volume: the number of applications, messages and referrals you send in a week. Stage 2 is the reply rate: the share of applications that get any answer from a person, a call from a recruiter or an invitation to a screening call, the short first call in which a recruiter checks the basics (availability, salary, motivation). Stage 3 is the interview rate: the share of replies that turn into a real interview. Stage 4 is the offer rate: the share of interviews, or of final rounds, that end with an offer.

The point of splitting it is that the stages fail for different reasons. A low volume is a time and rhythm problem. A low reply rate points at targeting, the CV or the channel. Replies that never become interviews point at the screening call or the fit with the posting. Interviews that never become offers point at the interview itself. Rewriting your CV will not help if your problem is at stage 4, and more interview practice will not help if nobody calls you back.

What do normal numbers look like?

There is no single benchmark, and anyone who quotes one without a source is guessing. What exists are a few dated measurements, each from a different population, which give an order of magnitude rather than a target.

The best general one comes from the New York Fed’s Survey of Consumer Expectations, analysed by Jason Faberman and colleagues (IZA discussion paper, 2017). Pooling the October 2013–2015 waves, unemployed job seekers (N = 163) had sent a mean of 8.14 applications in the last four weeks and received a mean of 1.261 employer contacts and 0.373 offers. Employed people who were looking (N = 508) sent fewer, 4.63, and received more: 1.874 contacts and 0.425 offers. Our own calculation from that table: about one offer per 22 applications for the unemployed (8.14 ÷ 0.373 = 21.8), about one per 11 for employed seekers (4.63 ÷ 0.425 = 10.9). These offers include a few the person never applied for (0.043 of the 0.373). Contacts include more of those: leaving out unsolicited contacts, the unemployed received 0.802 contacts for 8.14 applications, about 10 replies per 100 applications (0.802 ÷ 8.14 = 0.099).

Tracker data from job seekers points the same way. Huntr, a job application tracker, analysed 1.7 million applications logged by its users in 2025: applications sent with a CV tailored to the posting reached the interview stage or further 5.8 % of the time, against 3.73 % for untailored ones, and 18 % of successful seekers needed more than 100 applications before an offer. These are self-reported logs from one tool’s users, and the tailoring gap is a correlation, not an experiment. On the employer side in France, the Apec’s quarterly follow-up surveys of recruiters who posted managerial jobs with it found a mean of 37 applications per posting in 2016, with 22 % of postings getting fewer than 10 and 7 % more than 100.

Step 1: count your own funnel for the last eight weeks

Write four numbers for the last eight weeks: applications sent (including direct messages and referrals), replies of any kind, interviews, offers. Then divide each number by the one before it. Eight weeks is long enough to have more than a handful of applications and short enough that the market and your CV have not changed much.

An illustrative example, not a real person: 40 applications, 1 reply, 1 interview, 0 offers. The reply rate is 1 ÷ 40 = 2.5 %, under the one-in-twenty line; one interview is too few to judge the later stages. The stage to work on is stage 2, the reply rate, because 39 of the 40 applications went unanswered. A second example: 12 applications, 4 replies, 4 interviews, 0 offers. Here the reply rate is good (33 %) and the volume is low, about 1.5 a week, and four interviews without an offer start to point at the interview stage.

Which stage fails tells you the fix

Read the table from the first stage that looks weak, and fix only that one for the next four weeks. The thresholds are rough lines derived from the figures above, not norms: with fewer than about 20 applications, or fewer than three interviews, wait for more data before you conclude anything about a later stage.

Is it a matter of effort?

Partly, and the research says so plainly. In a 2001 meta-analysis of 73 studies, Ruth Kanfer, Connie Wanberg and Tracy Kantrowitz pooled the 28 studies (7,980 people) that linked search to its results, and found that job search behaviour was positively related to finding employment (corrected correlation .21, on a scale where 0 means no link and 1 a perfect one) and to the number of offers (.28), and negatively to the length of unemployment (−.14). Those are modest correlations: effort matters, but it explains a small part of who finds work, and the size of the links changed with how studies measured search: as general effort, or as intensity, meaning how often people did specific search activities.

So a low volume at stage 1 is worth fixing first, because it is the stage most under your control. But if your volume is already steady and the reply rate is low, sending twice as many of the same applications mostly doubles the number of silences. That is the moment to change what you send and where, not how much: the guide “Visualization or networking for a job hunt?” explains why messages to people you already know deserve a place next to cold applications, and the guide “A job that fits you in 30 days?” gives the weekly rhythm.

What structured help adds

A job search does not have to be a solo project. In a 2014 meta-analysis in Psychological Bulletin, Songqi Liu, Jason Huang and Mo Wang summarised 47 job search programmes evaluated by experiment or quasi-experiment: participants had 2.67 times the odds of finding a job compared with control groups. Programmes worked better when they taught search skills, improved self-presentation, built self-efficacy, encouraged proactivity, set goals and enlisted social support. In the authors’ words, job search interventions “effectively promoted employment only when both skill development and motivation enhancement were included.”

The practical reading: a workshop or a coach who only cheers you on, or a course that only teaches CV formats, does half the job. In France, check with France Travail which workshops and support you can access; in the US, check with your local American Job Center. A friend who reviews your funnel numbers with you every two weeks covers part of the social support.

Where the night ritual fits, and what the funnel cannot show

Once you know your weak stage, the 2-minute night ritual can rehearse the one step that lives there: opening tomorrow’s search slot, sending the message to a former colleague, the first minute of the screening call. Picture your own steps, not the call with the offer. In Dream Architect, “Write my own moment” takes a few words in the present tense, and at One small step, near the end of the full ritual, the choice “Write one bold line” is your cue to write tomorrow’s action on paper, such as “tailor the CV for the two postings saved today”.

The funnel shows where your numbers drop. It does not show the market around them: a sector that is cutting jobs, a region with few postings, a season when hiring slows. If every stage looks reasonable and offers still do not come, the cause may be outside your process, and the guide “How to get job offers in a competitive job market” looks at that side. None of this promises a job by a date; it tells you where your next hour is best spent.

Questions people ask next