Why am I not getting interviews? Diagnose it by symptom

Start with the symptom

Five symptoms cover almost every case, and each one points at a different stage. Getting no answer at all is a different failure from being rejected in four minutes, and both are different from screens that never become interviews. Advice aimed at the wrong stage costs weeks.

How much of this is the 2026 market?

Some of it is, and knowing how much keeps you from rewriting a resume that was already fine. Greenhouse counted 244 applications per job in 2025, up 111% from about 115 in 2022, across 6,000+ companies and more than 640 million applications. Ashby reported more than 300 applications per hire in May 2026, three times the 2021 level, and estimated candidates are about 50% less likely to reach an interview than five years ago. The base rate moved, so a reply rate that would have been poor in 2021 can be ordinary now.

What the market does not explain is a reply rate of zero. Gem’s 8% first-screen pass rate is an average across a very large pool, so a run of fifty applications with no first conversation is unusual enough to investigate rather than absorb. The rest of this page is the investigation.

Symptom 1: applied to 100 jobs, no responses at all

Zero is the one number that is almost never a market effect. Two causes produce it. The first is extraction: an applicant tracking system reads text out of your file before a person sees anything, and two-column layouts, headers, text boxes and tables can scramble or drop what they hold. Your record then looks empty when a recruiter searches, and nothing tells you it happened. The second is targeting: you are applying to roles whose title, level or location does not match what your resume evidences.

The 10-minute test. Paste your resume into the ATS extraction preview and read the text it returns. Check four things: your email and phone survived, the section headings are recognisable, the dates sit next to the right employers, and your job titles are not interleaved with your skills. Then open five postings you applied to and write your resume title line beside each posting title. If your line says Customer Success Lead, APAC and every posting says Account Manager, the search that a recruiter runs will not return you.

The fix. Repair the layout first, because everything else depends on it, then tailor the title and the summary per posting. Huntr recorded 4.23% interviews per application on tailored resumes in Q1 2026 against 2.07% on untailored ones, from 139,927 applications. The mechanics of both steps are covered in how to pass ATS filters and where resume keywords come from.

Volume belongs at the end of the list. Huntr’s Q1 2026 data has job seekers who sent 11 to 20 applications interviewing at 9.25% per application, and those past 100 at 2.58%. That is a correlation, since the two groups differ in more ways than volume, but it does say that sending more untested applications is a weak answer. The arithmetic of how many to send is in how many jobs to apply to.

Symptom 2: rejected within minutes of applying

A rejection that arrives in four minutes was decided by the form. Employers can attach knockout questions to an application (work authorisation, willingness to relocate, minimum years of experience, a pay figure, a required licence) and set an answer to end the application on the spot. Hard filters do the same job more quietly: a location radius, a salary ceiling, a degree requirement. In the Harvard Business School and Accenture Hidden Workers research, more than 90% of employers said they use their recruiting software to filter or rank candidates at first screening, and 88% agreed that qualified high-skills candidates are screened out because they do not match the exact criteria in the job description. For middle-skills roles the figure was 94%.

The 10-minute test. Reopen the application form for a role that rejected you quickly and reread your own answers. People fail these fields by being careless rather than unqualified: a notice period entered in weeks where the form wanted a date, a salary box filled with a stretch number, a relocation question answered no because the role looked remote. Then look at your tracker and count applications per company. Huntr measured 6.07% interviews per application for people who applied once at a company in Q1 2026, and 1.91% for those who sent eight or more.

The fix. Answer the fields as your situation actually qualifies, and stop applying where a field rules you out: a role requiring authorisation you do not hold is a no before anyone reads a word. Choosing postings you can pass is the subject of targeting the right jobs. If the filter you keep failing is a gap in your history, HBS found 48% of employers filtering out middle-skills candidates and 43% filtering out high-skills candidates for a resume gap, which is worth pre-empting the way explaining an employment gap sets out.

Does the ATS reject most resumes automatically?

The claim that applicant tracking systems reject 75% of resumes, or that 80% never reach a human, has been repeated for years without a study behind it. We could not trace either number to a published method or sample, so neither belongs in a diagnosis. The evidence that does hold up is the HBS Hidden Workers survey above: filtering is real, it is configured by employers, and it works on exact criteria rather than on an overall score.

The practical difference matters. If software scored your resume, you would need a better score. Since an employer set a filter, you need to match the criterion, or to stop applying to postings where you cannot. And the failure that costs most is not a filter at all: it is text that never made it out of your file, which is why the extraction check comes before everything else.

Symptom 3: recruiter screens, then nothing

Getting the call means your resume worked. The screen is a different test, and it is the narrowest stage of the process: Ashby’s 2026 operations benchmarks put pass-through at early stages such as recruiter screens at about 35%, from 54 million applications and 93,000 jobs. Three things end screens. Your pitch answers a question the posting did not ask. Your pay expectation lands outside the band. Your availability or location turns out to be more complicated than the form suggested.

The 10-minute test. Record your answer to the opening question on your phone, keep it to 60 seconds, and play it back against the top three requirements in the posting. Count how many of the three you actually touched. Most recordings cover a career history and none of the three.

The fix. Rewrite the pitch as three claims, one per requirement, each with a number or a named outcome attached, and decide your pay range and notice period before the call. The same requirement-by-requirement reading is how tailoring works on paper, covered in tailoring a resume to a job description. One more lever at this stage: Ashby found 52% of referred candidates pass the first screen against 35% overall, and Gem put referral conversion at 11 times the inbound rate.

Symptom 4: interviews but no offers

At this stage you are being compared, no longer screened. Gem’s benchmarks have 8% of applicants passing the first screen and 0.5% receiving an offer, with interviews per hire up 33% since 2021, so later rounds now involve more people and more of them are close on paper. Losing at this stage means someone answered the same questions with more specific evidence.

The 10-minute test. Within an hour of each round, write two lines: the question you answered worst, and the topic the interviewer returned to twice. After three interviews the pattern is usually obvious, and it is usually the same missing example.

The fix. Build two or three worked examples that carry numbers, dates and your own decisions, and reuse them. Then ask the recruiter what the panel wanted more of, once, in one sentence. A rejection is the cheapest feedback available in a process where most candidates get none.

Symptom 5: silence, ghosting and postings that were never real

Silence is a stage of its own in 2026, and some of it has nothing to do with you. Greenhouse surveyed 2,950 job seekers in the US, UK, Germany, Australia and Ireland in May 2026: among candidates who completed an AI interview, 28% moved forward, 13% were formally rejected, and 51% never heard back. In Greenhouse’s July 2025 workforce survey of 2,200 people, 63% said they had been left without news after interviews. And in Huntr’s Q1 2026 survey, 520 of 559 respondents (93%) said they had applied to a posting they believed was fake or never meant to be filled, or suspected it.

The 10-minute test. Before tailoring anything, check the posting: how old it is, whether the same text has been reposted every few weeks, whether the company careers page carries it at all. Then check your own clock. Ashby found candidates who never get an interview are archived in a median of about six days, while interviewed candidates average 21 days, so two weeks of silence after an application is usually a decision that nobody sent.

The fix. Follow up once, a week after applying and a week after an interview, then let it sit and keep your weekly count steady. Spending the freed time on referrals pays better than spending it on follow-ups, given the referral numbers above.

A two-week plan to test the diagnosis

Diagnosis only helps if you change one variable at a time. Two weeks is enough to see whether the first conversation rate moves, and it is short enough to keep.

• Day 1. Run the extraction check and fix whatever did not survive it. Nothing else matters until contact details, headings and dates come back intact.
• Days 2 to 5. Send 8 to 10 applications, one per company, each with the title line and summary rewritten against the posting. Log the date, the company, the title you used and the form fields that gave you pause.
• Days 6 to 10. Add two referral attempts. A referred application passes the first screen at 52% against 35% overall in Ashby’s data, which is worth more than ten extra cold applications.
• Days 11 to 14. Count first responses instead of applications. One or two first conversations from ten targeted applications is a working pipeline at 2026 rates. Zero, with clean extraction and honest targeting, points at the roles you are choosing rather than the documents.

If you would rather run this as a standing system than as a fortnight of effort, the guide to getting a job in 2026 covers the tracking side.

Frequently asked questions

Sources

1. Huntr, Job Search Trends Q1 2026. 139,927 applications from 25,635 job seekers, plus a 593-person survey.
2. Gem, key takeaways from the 2026 Recruiting Benchmarks report, December 2025. 165M+ applications, 1.2M hires.
3. Greenhouse, Hiring Benchmarks, May 2026, with coverage by RecTech Media, 9 May 2026.
4. Ashby, Recruiting Operations Benchmarks, 2026. 54M+ applications, 93K jobs, January 2021 to March 2026.
5. Ashby, press release on application volume, 7 May 2026.
6. Greenhouse, AI interview survey, 1 May 2026. 2,950 job seekers in the US, UK, Germany, Australia and Ireland.
7. Greenhouse, 2025 Workforce and Hiring Report, 17 July 2025. 2,200 respondents.
8. Harvard Business School and Accenture, Hidden Workers: Untapped Talent, September 2021. 8,720 workers and 2,275 executives in the US, UK and Germany.