The Hiring Funnel Leak Map: Find Where You Lose Candidates Before You Spend More

When there aren't enough applicants, the first thing everybody says is "we need more budget." The hiring manager says it, the recruiter says it, sometimes the agency says it because more budget is easier to sell. And I get it, it's the most obvious lever. But most of the time the problem isn't the top of the funnel. It's a leak somewhere further down, and more budget just pours more people into the same hole.

Think about it like a bucket with a crack in it. You can keep filling it faster, and yes, the water level goes up a little. Or you can find the crack.

Why more budget is usually the wrong fix

If 100 people click your job ad and 95 of them leave on the application form, doubling your budget gets you 200 clicks and 190 people leaving on the form. You paid twice as much for the same problem. The cost per hire stays bad, or gets worse, because the extra clicks you buy are usually a bit less relevant than the first ones.

The thing is, most companies can't even tell you where the leak is. They know how much they spent and how many people they hired, and everything in between is a black box. Somebody looks at the number of applicants, decides it's low, and the only question that gets asked is "how much more do we need?"

So the move is simple. Treat hiring like a marketing funnel. Measure every step, find the biggest leak, fix that one first, and only then decide if you need more money.

The 9 steps

Every arrow in that diagram is a spot where people drop out. Some drop-off is fine, you don't want everyone to apply. A leak is when you lose the good people at a step, for a reason you could fix. That's what we're looking for.

A quick word on benchmarks, because people always ask "what's normal?" I'm not going to make up percentages. The one public number I trust is from Appcast's 2025 Recruitment Marketing Benchmark Report, which looked at 2024 job ad data from more than 1,300 U.S. employers: the overall apply rate, so applications per job ad click, ended 2024 at 6.1%. That's an average across all kinds of jobs, and it's U.S. data, so use it as a rough reference, not a target. Appcast's 2026 report added benchmarks for the steps after the click (screen, interview, offer, hire), so if you want outside numbers for the bottom of the funnel, that's the place to look. For everything else, your best benchmark is your own other vacancies.

Here's each step, what to measure, what the leak usually looks like and what to do about it.

Impressions. The very top. You measure reach, frequency and CPM. This is the one step where budget really is the fix, but only if the reach is low. If you're reaching plenty of people and nothing happens after that, more reach won't help. Also check your targeting isn't so narrow that the same few thousand people see your ad ten times.

Impression → click. You measure click-through rate. Low CTR means people see the ad and don't care. Usually it's a job title nobody searches for, an ad that looks like every other ad, or a message written for job seekers when the people seeing it are scrolling. The fix is the creative, not the budget. Read Stop Writing Job Ads for Job Seekers, and if you think the title is the problem, test 5 job titles. If the ad has run for weeks and the CTR is slowly dropping, that's creative fatigue, so make a new one.

Click → lands on the ad. This one gets skipped all the time. You compare clicks in your ad platform with page views on your vacancy page. If those numbers are far apart, people clicked and never saw the page. Slow loading, a redirect chain through three job board pages, a broken link, or the job was already closed. You recognize it by a big gap between "clicks" and "sessions." Fix the page speed and the link, and check it on your phone.

Lands → starts applying. You measure the share of visitors that click "apply" or start the form. Low here means the page itself doesn't sell the job. No salary, a wall of duties, the first lines say nothing about what the job actually is. Do the 10-second mobile test, then score the ad with the Job Ad Scorecard. Most of the time it's fixable in an afternoon. Cut the clichés (The Cliché Tax) and swap the duties list for a day in the job.

Starts → finishes applying. Application completion rate. If lots of people start and few finish, the form is the problem. An account, a cover letter, re-typing your CV after uploading it, 25 fields on a phone. This is one of the most common leaks I'd check first, because it's invisible unless you measure it. Fix it with how to get your application under 5 minutes. And if you're worried that an easier form lets in junk, read Your Next Applicant Is a Bot for friction that filters instead of scares.

Finishes → first contact. You measure two things: the share of applicants who get contacted, and how long it takes (speed to lead). This is where a lot of good people quietly disappear. They applied on Monday, you called on Friday, and by then they have two other conversations going. You recognize it when candidates don't pick up or say "oh, I already found something." The fix is a rule, for example every applicant hears from a human within one working day, and someone who owns it.

First contact → interview. Measure how many contacted candidates actually sit an interview, and how many no-shows you get. Leaks here are scheduling back and forth for a week, an interview that's three weeks out, or a phone screen that feels like an interrogation. Let candidates pick a slot themselves, keep the gap short, and tell them what to expect.

Interview → offer. The offer rate. If you interview a lot of people and offer almost nobody, it's usually not the candidates. It's a mismatch between what the ad promised and what the hiring manager actually wants, or a list of must-haves nobody can meet. That's an intake problem. Use the three must-haves script with the hiring manager before you touch the ad again.

Offer → hire. Offer acceptance rate. When people say no at the very end, check three things. Was the salary a surprise? Was the process so long they took another job? Did something in the interviews not match the ad? Salary surprises are the easiest one to kill, just put the range in the ad.

How to find your biggest leak in one afternoon

You don't need a data team for this. Pick one vacancy that's been open for a while, ideally one with at least a few weeks of traffic. Pull the numbers for each of the 9 steps from your ad platform, your analytics and your ATS, and put them in the sheet below. Then work out the conversion from each step to the next.

Now look for the step where you lose the most compared to your own other vacancies, or compared to the Appcast reference for click to apply. If you have no comparison at all, start at the top and fix the first step that clearly looks broken, because a leak higher up hides everything below it.

Fix one thing. Only one, otherwise you won't know what worked. Run it for two weeks, fill in the sheet again, and see if the leak got smaller. Then go to the next one. Only when every step looks healthy and you still need more people, then yes, raise the budget. That's when the money actually works.

The sauce: the leak diagnosis sheet

And here's the simple tracking setup so you actually have these numbers next time.

1. UTM TAGS ON EVERY AD LINK
   ?utm_source=[platform]&utm_medium=paid&utm_campaign=[job-slug]&utm_content=[ad-version]
   One campaign name per vacancy. One content value per ad version.

2. THREE EVENTS ON YOUR VACANCY PAGE (in your analytics tool)
   vacancy_view    → the page loaded
   apply_start     → someone clicked apply or typed in the first field
   apply_submit    → the application was sent (thank-you page or form success)

3. FIVE DATES IN YOUR ATS FOR EVERY CANDIDATE
   applied_at, first_contact_at, interview_at, offer_at, hired_at
   If your ATS has stages, make sure each stage change is saved with a date.

4. ONE WEEKLY ROW PER VACANCY IN A SHEET
   Copy the numbers in every Monday. Takes 10 minutes.

week,vacancy,impressions,clicks,page_views,apply_starts,apply_finishes,contacted,median_hours_to_contact,interviews,offers,hires,spend
2026-W45,icu-night-nurse,,,,,,,,,,,

Add one formula column per step (step ÷ previous step) and color the lowest one red. That red cell is your leak.

Frequently asked questions

What is a funnel leak in hiring?

It's the step in your hiring funnel where you lose the most good people, for a reason you could fix. Not every drop-off is a leak. Losing people who were never a fit is fine. Losing qualified people because your form takes 20 minutes is a leak.

What's a good apply rate?

Appcast's 2025 benchmark report put the overall U.S. apply rate at 6.1% at the end of 2024, measured as applications per job ad click. That's an average across all jobs, so your role and market can be way off from it. Your own vacancies are the better comparison.

Should I ever just raise the budget?

Yes, when the steps after the click look healthy and you simply don't have enough people coming in. At that point more budget works the way you hoped. Before that, it mostly makes the leak more expensive.

How long before I can see if a fix worked?

Give it about two weeks and enough traffic that the numbers aren't just a few people. With small volumes, treat the result as a direction, not proof.

Sources

• Appcast: The 2025 Recruitment Marketing Benchmark Report is here (February 11, 2025), for the 6.1% overall apply rate at the end of 2024
• PR Newswire: Appcast releases 10th annual Recruitment Marketing Benchmark Report with new candidate disposition data (February 17, 2026), for the click-to-hire benchmarks in the 2026 report