Your Developer Job Ad Gets No Applicants: A 15-Point Diagnostic Before Reposting
Fifteen checks in the order a candidate meets your ad: findability, the ten-second scan, credibility, process and market. Each with a symptom, a test and a fix.
An ad for a developer role goes live, a week passes, and nothing comes in: no applications, or three that have nothing to do with the job. The usual reaction is to repost the same text on a second board and pay again. This diagnostic is for the recruiter, agency or hiring manager who would rather find out why first. It runs fifteen checks in the order a candidate meets your ad: can they find it, do they stay past the first ten seconds, do they believe it, and can they actually apply. Each check has a symptom, a test you can run in minutes and a fix. Where a number appears, it comes from a dated public source linked in the text; everything else is mechanism, not statistics. skillfind.tech, which publishes this page, is a free IT skill board and one of the places you can repost the corrected ad, and it is held to the same checks.
Which symptom do you have
Before the checks, look at the numbers your board gives you. Many boards show views and applications per ad, and some show clicks on the apply button. The pattern tells you where to start.
- Few views. The ad is not being found. Start with findability (checks 1 to 4), then look at ad age and refresh (check 14) and at niche skills (check 15).
- Views but no applications. People open the ad and leave. Start with the ten-second scan (checks 5 to 8), then credibility (checks 9 to 12) and the application path (check 13).
- Applications, but from the wrong people. The ad attracts, but not the right candidates. Start with the title and the tags (checks 1 and 2), the level (check 10) and the pay line (check 5), then the market (check 15).
Calibrate "no applicants" by level before you panic. On No Fluff Jobs, where every ad carries a range, the average senior ad drew 19 applications in 2025 and the average mid ad 28, down from 31 and 48 a year earlier (2025 report, via Antyweb). Junior ads, by contrast, routinely drew more than a hundred in 2024 (2024 summary, via android.com.pl). A senior ad with a handful of applications in its first week is close to normal; a junior ad with none is almost certainly not being found.
If the board shows you nothing, run a proxy: search the board the way a candidate would, open your own ad on a phone, and time how long it takes to apply. The checks below tell you what to look for.
Findability: can the right candidate find the ad
1. The title is not what candidates search for
- Symptom: few views, and the ad does not appear when you search the board for the technology.
- Why it happens: candidates search and filter by technology, role and level. An internal title ("Software Engineer II", "Member of Technical Staff"), a creative one ("Code Wizard"), a product name nobody outside the company knows, or an abbreviation used only in your organisation hides the ad from search and from the eye.
- Test: type into the board's search what a candidate would type: "Python", "React developer", "senior Java". Does your ad appear, and does the title say in one glance the technology, the role and the level?
- Fix: "Senior Java Developer (Spring, Kubernetes)" or "Junior Python Developer, hybrid, Kraków". Put the main technology in the title and the level in front of the role, and keep the whole line under sixty characters, because some boards and most search results cut it there. If the internal title must survive, it goes in the description, not in the headline.
3. The board does not have your audience
- Symptom: few views everywhere, or views only from people who have nothing to do with the role.
- Why it happens: a general classifieds site is not where a senior Rust developer looks, a developers-only board is not where a product owner looks, and a professional network reaches mostly people who are not looking. No single channel dominates. In Bulldogjob's Polish IT Community Report 2025, based on 4,331 respondents, people most often learned about their current employer from friends (27 per cent). General job portals (19.7 per cent), IT job portals (18.5 per cent), headhunters (14.5 per cent) and social media (10.8 per cent) followed. Every board also has a language habit: in Polish IT an English ad is common, but a board whose candidates expect Polish will still read an English-only ad as a signal about the team.
- Test: on the board, filter for your main skill and level. How many competing ads are there, how old are they, and do they look like yours? None means no audience; hundreds means you are one of hundreds.
- Fix: match the board to the role. Our comparison of Polish IT boards, Where to Post IT Jobs in Poland, lists what is free and roughly what the paid tiers cost. A workable mix is one board where candidates filter by skill, one professional network and one free board.
4. The location and mode fields contradict the text
- Symptom: the ad is filtered out by city or remote filters even though the role fits.
- Why it happens: candidates filter by city and by remote, hybrid or office before they read a word. If the field says Warsaw and the text says "fully remote", the remote filter never shows the ad. "Poland" typed as a city, a headquarters address on a remote role, "hybrid" without naming the office city: each one loses a filter.
- Test: set the board's filters the way a remote candidate and a local candidate would. Does the ad appear in every case that applies?
- Fix: fill the fields to match the truth: the city where the office is, the mode as the board defines it, and several locations if the board allows them. Then say in the first lines how many days in the office, and whether remote means anywhere in Poland or anywhere at all.
The ten-second scan: do they stay
Candidates read ads the way you read CVs: a glance at the top, a decision in seconds. The next four checks are what they look for in that glance.
5. There is no pay line
- Symptom: views but almost no applications, and the candidates who do apply ask about pay in their first message.
- Why it happens: the pay line is usually the first thing a developer looks for and often the first filter they use. Two Polish IT boards, No Fluff Jobs and Solid.jobs, do not publish an ad without a range at all. Just Join IT and theprotocol.it let candidates show only ads with pay. The chief executive of Just Join IT said in a September 2026 interview that ads with a range bring two to three times more CVs than ads without one. That is the board's own assessment. Candidates say the same from their side. In No Fluff Jobs' 2023 candidate experience survey of 1,216 Polish IT candidates, 45.5 per cent said they do not send a CV to an ad without a range, rising to 49.5 per cent among mid-level and 64 per cent among senior candidates. In the same survey, 80 per cent named pay as the thing most often missing from ads. Since 24 December 2025 Polish law has required an employer recruiting for an employment contract to tell the candidate the starting pay or its range, in the ad or, if not there, before the interview (summary on HRappka, not legal advice). The number therefore exists anyway. An ad without it reads as "we will pay as little as you accept".
- Test: can a stranger tell in ten seconds what the role pays, in which contract form and per what period?
- Fix: write the range the way the market reads it: B2B net per month, or per hour if that is how you pay, and UoP gross per month, both if you offer both, with the currency. A real range beats a wide one; a range whose top is double its bottom tells candidates you have not decided. In Grant Thornton's August 2026 pay transparency report, covering 292,054 ads across all sectors on Pracuj.pl, Praca.pl and Rocketjobs.pl from spring 2026, only about a quarter stated pay at all. Of the published ranges, 15 per cent were wider than 80 per cent, which the authors read as some employers treating the published pay as no more than a rough indication. If you truly cannot publish, say what you will tell them and when. Silence is the worst option.
6. The contract form is unclear
- Symptom: candidates apply and drop out at the first call when they hear the contract form, or seniors skip the ad altogether.
- Why it happens: in Polish IT, B2B (a contract with the developer's own business) and UoP (umowa o pracę, an employment contract) are different offers: different net pay for the same cost to you, different leave, notice periods and benefits. The convention, as Just Join IT's July 2026 explainer puts it, is B2B quoted net on the invoice and UoP quoted gross. B2B is the market's default form. It appeared in 63.3 per cent of IT ads on inhire.io in the third quarter of 2025 (via ITwiz) and in about 72 per cent of mid and senior ads on No Fluff Jobs in 2025 (via Antyweb). Among junior ads on the same board, its share fell to 37.9 per cent. The specialists themselves split almost evenly: 48.7 per cent were on B2B and 45.5 per cent on an employment contract in Just Join IT's 2025 survey of over 4,400 IT professionals. Whichever single form you offer, about half the pool works on the other one. Many seniors want B2B, many juniors and people with mortgages want UoP. "Attractive terms" tells neither group anything, and a B2B rate quoted next to a UoP salary without labels confuses both.
- Test: does the ad say which contract forms are available, and does the pay line match each form?
- Fix: name the forms and the pay for each in one line, for example "B2B: range net per month; UoP: range gross per month". If you offer only one form, say so plainly. Candidates prefer a clear no to a surprise at the first call.
7. Remote, hybrid or office is not clear
- Symptom: candidates drop out when they learn the office days, or remote candidates never apply.
- Why it happens: "hybrid" means anything from one day a month to four days a week, and "remote" sometimes means remote within Poland, sometimes remote within commuting distance "for occasional meetings". Candidates have been burned before and now assume the worst reading. Remote is not a perk in Polish IT but the norm. In Bulldogjob's 2025 survey, 59.9 per cent of respondents worked remotely every day and only 8.6 per cent never did, and the possibility of remote work was the second most important thing at work (21 per cent) after money (25.9 per cent). On the supply side hybrid now leads: in inhire.io's third-quarter 2025 data (via ITwiz), 58.8 per cent of ads were hybrid, 30.7 per cent remote and 10.3 per cent office only. An office-only ad is therefore one ad in ten, read by candidates of whom fewer than one in ten work that way.
- Test: can a candidate say, from the ad alone, how many days a month they would spend in which office, and from which countries they may work?
- Fix: one sentence with numbers: "Hybrid, two days a week in the Wrocław office" or "Remote from anywhere in Poland, one team day in Warsaw per quarter, travel paid". If the mode is negotiable, say that it is negotiable.
8. The role is buried under company copy
- Symptom: views, a short time on the page, few applications.
- Why it happens: the first screen on a phone holds three or four short paragraphs. If they are about the company's awards, mission and growth, the candidate has to scroll to find out what the job is and what it pays. Many do not.
- Test: open the ad on a phone. What do the first three paragraphs say? If "we" appears more often than "you", rewrite them.
- Fix: first the facts (role, level, stack, mode, city, pay, contract), then three lines about the product and the team, then the requirements, then the company. The company section can stay; it simply goes last.
Credibility: do they believe it
9. Impossible or inflated requirements
- Symptom: views and no applications, or applications only from people who ignore requirements.
- Why it happens: twelve must-have technologies, five years of experience in a framework that is three years old, senior skills under a junior title, or three jobs in one (backend, frontend, DevOps and some data engineering on the side). Experienced candidates read these as a company that does not know what it needs, and they select themselves out. Juniors, who are told to apply anyway, apply; seniors do not.
- Test: for each requirement ask two questions: will we test for this in the interview, and would we reject a good candidate without it? If either answer is no, it is not a must-have.
- Fix: three to five must-haves, the rest labelled nice-to-have or removed. Replace years with what the person should be able to do. Check how old a technology is before you write a number of years next to it.
10. The level does not match the requirements or the pay
- Symptom: applications from the wrong level: seniors ignore a "senior" ad, juniors do not apply to a "junior" ad.
- Why it happens: candidates infer the real level from three signals, the title, the requirements and the pay, and distrust the ad when the three disagree. A "Junior" ad asking for three years of production experience at junior pay is a mid ad at junior pay. "Regular", common in Polish ads, means Mid, which a foreign candidate may not know; the filters on two major boards run Intern, Junior, Mid, Senior and then Team Leader or Lead (Just Join IT, Bulldogjob), and "Regular" is not among them. The market is also top-heavy: almost 60 per cent of open recruitments in No Fluff Jobs' 2025 data were senior and about 35 per cent mid, with juniors at about 5.3 per cent of ads (via Antyweb). A mid ad dressed up as senior therefore joins the most crowded shelf, and a junior ad is one of very few.
- Test: do the title level, the years or responsibilities asked for, and the pay range all describe the same person?
- Fix: pick the level by what the person will own unsupervised, then set the pay and the requirements to that level. On skill-tagged boards set each skill's level honestly: a Senior tag on every skill of a mid role is a filter against the people you want.
11. The ad is anonymous
- Symptom: views and no applications, candidates asking who the client is before they apply, or an ad that is simply skipped.
- Why it happens: agencies post for unnamed clients, and some companies hide behind "a leading company in its sector". Developers read anonymity as risk: a company they would avoid, a role that may not exist, a CV forwarded without consent. A named recruiter and a described product change that reading.
- Test: does the ad say what the product does, who uses it, how big the team is, what the stack is and who the reader will talk to? Would you apply to it without knowing the company's name?
- Fix: if you cannot name the client, describe everything else: industry, product type, team size, stack, office city, the process and a named recruiter with a direct address. Say why it is confidential. Check the board's rules too: skillfind.tech's terms require the poster to be a business acting in its own name and say nothing against recruiting for a client, so an agency posts under its own name and describes the client's role.
12. No stack, no context, or stale context
- Symptom: views, few applications, and the candidates who apply have not understood the role.
- Why it happens: "modern technologies", "dynamic environment" and "exciting projects" carry no information. Candidates want to know what they would build, for whom, with what, and what will change in the next year. A stack list that ends in a technology nobody has used for five years signals an ad copied from an old one. In No Fluff Jobs' 2023 candidate experience survey, 70.4 per cent of candidates wanted the technologies used to be listed in the ad.
- Test: cover the company name. Could a developer tell from the text what they would be working on next month?
- Fix: three lines: what the team builds and for whom, the stack as it is today (be honest about legacy), and what the next year holds, such as a migration, a new product or scaling. Date-check the stack list every time you repost.
Process, timing and market: can they apply, and would anyone
13. Applying is harder than it needs to be
- Symptom: clicks on the apply button, few completed applications.
- Why it happens: an external link that demands an account, a twenty-field form that asks for everything already in the CV, a mandatory cover letter, a broken link, a form that does not work on a phone, or "send your CV to recruitment@company" with no name attached. Every step loses people, and developers are among the least patient candidates on the market, because many of them rarely need to apply at all. In a 2022 Greenhouse survey of job seekers in the US, the UK, Ireland and Germany, more than 70 per cent said they would not complete an application that takes longer than 15 minutes, and more than half expected to hear back within a week.
- Test: apply to your own ad from a phone, in a browser where you are not logged in, and time it. More than a few minutes, or more than one page, is the problem.
- Fix: a CV or a profile link plus two questions at most, no account, no cover letter, and a confirmation message that names the next step and a date. If your applicant tracking system cannot do that, let candidates apply through the board's own form and move them into the system yourself.
14. The process and the timing work against you
- Symptom: applications come in and then the candidates go silent, or the ad is old and the views have stopped.
- Why it happens: candidates drop out when they hear about five interview stages and an unpaid weekend task, when the first reply takes two weeks, or when the ad gives no idea of the process at all. Ads also age: new ads push older ones down the list, and the boards sell bumps, counted per package, precisely to push an ad back up. Most paid tiers run for about a month (Just Join IT's help centre gives 30 days as standard), and a skillfind.tech posting stays live for 60 days. The Polish benchmark for speed is not flattering. In eRecruiter's recruitment KPI report for the second half of 2025, built on 10.5 million applications, the average time from a CV being sent to a recruiter opening it was 9 days, and the average time to an offer 20 days. In the same report, candidate satisfaction with recruitment, measured as a net promoter score, fell to 37 points, its lowest ever. One in five candidates in Greenhouse's 2024 candidate experience report, the vendor's own survey, had rejected an offer because of a poor interview experience.
- Test: how many steps and how many hours does your process take, and does the ad say so? When did the ad last move up the list? How fast did you answer the last applicant?
- Fix: describe the process in the ad in three steps with the expected time for each. Reply within two working days, even if only to say when the decision will come. Cap any task at a few hours or pay for it. Refresh or repost when the ad stops moving, and change something when you do. Mind the calendar as well. No Fluff Jobs counted 60 per cent more applicants in January 2025 than in December 2024 (via Antyweb). Grant Thornton's market monitor describes April to July as the busiest months for ads, with a drop in August (July 2025 report), and notes that January usually brings more ads than December (December 2025 report). An ad posted in mid-December is therefore judged in the slow weeks before the January rush.
15. The market says no
- Symptom: everything above checks out and still nothing comes.
- Why it happens: the pay is below what competing ads for the same stack and level offer in that city, the stack is rare in Poland, the role is office-only in a city where the specialists are not, the ad requires fluent Polish for a role that could be done in English, or the level you want is senior and the pay is mid. An ad cannot fix an offer, and pay is where to look first: earnings would motivate a job change for 88.7 per cent of Polish IT specialists in Bulldogjob's 2025 survey, far ahead of flexibility at 39.5 per cent.
- Test: open ten competing ads for the same skill and level on boards that publish ranges, and read their pay, mode and requirements. Where does yours sit? Remember that there may be more ads to compete with than you think: No Fluff Jobs counted 44 per cent more ads in 2025 than in 2024, and the average employer there posted 17 ads instead of 10 (2025 report, via Antyweb).
- Fix: change the offer, not the wording: raise the range, open remote or hybrid, hire a mid and train, drop the language requirement, split the role in two, or accept a longer search and say so to the hiring manager. Then repost, because the old ad has already told the market what it was.
Fix, then repost
Fix everything that failed, not the one check you like. Change the title and the first lines, because candidates who scrolled past the old ad will recognise it. Then repost or refresh, and give it a week before you judge it.
Reposting costs money on most boards, so put the corrected ad on a free channel as well. One option is to add a posting for free on skillfind.tech. Tag one to ten skills with a level each, write the pay line in the remuneration section as B2B, employment or both, pick the mode and the city, and the posting is live for 60 days in English and Polish. It is a small board and it will not fix a market problem; it is simply one more channel for a corrected ad.
The fifteen checks in one list
- Title: technology, role and level, under sixty characters.
- Tags: the must-haves with their real levels plus two or three nice-to-haves, nothing else.
- Board: one where candidates filter by your skill, plus a network and a free board.
- Fields: city, mode and location match the text.
- Pay line: a real range, with contract form, period and currency.
- Contract: B2B, UoP or both, with the pay for each.
- Mode: days in the office, which office, and from where remote work is allowed.
- First screen: facts first, company last.
- Requirements: three to five must-haves, abilities instead of years.
- Level: title, requirements and pay describe the same person.
- Identity: product, team, stack, process and a named recruiter, even when the client stays unnamed.
- Context: what the team builds, the stack today, the next year.
- Applying: a few minutes on a phone, no account, no cover letter.
- Process and age: steps and time in the ad, replies within two working days, a refresh when it stops moving.
- Market: pay, mode and requirements measured against ten competing ads.
This diagnostic was written in October 2026. If a board changes a rule we mention, tell us through the contact page and we will correct it.