Recruitment Funnel: Stages, Conversion Rates and How to Fix the Bottleneck

When a vacancy is stuck, the first thing I open is not the list of candidates. It is the funnel: how many people entered, how many reached each stage, and where they disappeared. In the course I call the recruitment funnel a powerful analytical tool for working with a vacancy, because it shows at once where the problem is.

This guide covers the stages, the conversion formula with a real vacancy, the reverse calculation — how many candidates you need for a given number of hires — and what to fix at the top, the middle and the bottom.

If you already have the numbers, jump to the free recruitment funnel calculator.

The idea comes from marketing and sales. At the entrance there are potential candidates, at the exit there are hires, and at every stage in between there are fewer people than at the one before. A funnel shows how many people move from one stage to the next.

Two things are worth keeping in mind.

The funnel is the company’s view of the process. The same process seen through the candidate’s eyes is a candidate journey map. I treat the funnel as a simplified version of that map: the map describes how one person goes through the stages, the funnel counts how many people got through each of them.

The candidate is making a big decision. For most people choosing a job is close to choosing a car or a flat, because the job is what pays for them. So the candidate moves through your stages much like a buyer of something expensive. That is why I teach the funnel together with sales techniques.

The basic funnel I use in the course, from top to bottom:

1. Candidates in the funnel — resumes, profiles and applicants added to the vacancy.
2. Responses — people who answered your contact.
3. Phone screening — candidates who turned out to be qualified, interested and ready to move.
4. Interview — candidates submitted for consideration and invited.
5. Final interview.
6. Job offer made.
7. Hire — the offer is accepted.

I also split it into three zones, because they are fixed with different tools:

• Top — where candidates come from: proactive sourcing, your own database in the ATS, advertising, mailings to a partner database, job boards, social referrals and employee referrals.
• Middle — phone screening, interview, final interview.
• Bottom — the offer and the decision.

Your stages do not have to match mine. One rule matters: every stage must be an event you can check — it either happened or it did not. “In progress” is not a stage.

Before analysing a funnel I ask the recruiter five questions. What percentage of candidates added to the funnel get in touch? What percentage do you invite to an interview after the phone screening? What percentage reach the final interview? What percentage get an offer after the final interview? What percentage accept it? And a sixth one: how many positions do you want to fill with this funnel?

Here is one real vacancy that I show in my courses, stage by stage:

• 301 candidates at the top of the funnel
• 68 phone interviews — 68 ÷ 301 = 22.6%
• 34 HR interviews — 34 ÷ 68 = 50.0%
• 20 recommended to the hiring manager — 20 ÷ 34 = 58.8%
• 15 passed the test task — 15 ÷ 20 = 75.0%
• 3 finalists — 3 ÷ 15 = 20.0%
• 1 hire — 1 ÷ 3 = 33.3%

Overall conversion: 1 ÷ 301 = 0.33%.

Some stage conversions have their own names, and they are worth tracking separately:

• Response rate — how many people answered the first contact (email, message or call).
• Interview show-up rate — came to the interview ÷ invited to the interview.
• Offer acceptance rate — accepted offers ÷ offers made.
• Channel quality — hires from a channel ÷ candidates from that channel.

The funnel is read from the top down, but it is planned from the bottom up. Start with the number of hires and go backwards:

Candidates needed at a stage = candidates needed at the next stage ÷ conversion rate between them

And always round up. You cannot hire half a person; round down and the plan falls short.

Take the vacancy above and suppose I need three hires for the same role:

• 3 hires ÷ 33.3% = 9 finalists
• 9 ÷ 20.0% = 45 candidates who pass the test task
• 45 ÷ 75.0% = 60 recommended to the hiring manager
• 60 ÷ 58.8% = 102 HR interviews
• 102 ÷ 50.0% = 204 phone interviews
• 204 ÷ 22.6% = 903 candidates at the top

Then turn the top number into weeks: divide it by how many new candidates a recruiter really adds per week. This is the capacity check: if one recruiter needs all of their time for this plan, the other vacancies wait.

I do not trust universal tables. A conversion rate depends on the role, on where the candidates came from and on how you defined the stage. Three comparisons tell you more:

1. With your own closed vacancies for a similar role. This is where planning numbers like my 300 at the top come from: we set them from experience and then accepted them as our standard.
2. Between recruiters on the same kind of vacancies.
3. Between sources — which channel gives hires, not only candidates.

One number of my own. Where I built the recruiting system, offer rejections were always below 5%. That is not the result of some brilliant selling at the last step. An offer should not be made to everyone: it is made to people who fit the vacancy and who will want to work there, and whether a person wants it is decided by how the vacancy was presented at every stage before the offer.

Start with the lowest stage conversion, but do not stop at it. In the vacancy above the narrowest step is test task → finalists, 20%. A step is a bottleneck when it is worse than the same step in your other vacancies, or when candidates leave it for a reason you could remove.

The usual bottleneck in each zone, as I describe them in the course:

• Top — nobody knows about you, and information about the company cannot be found.
• Middle — no sales techniques are used when contacting and interviewing candidates.
• Bottom — the job is not presented as a product together with the offer. Worse, there is no written job offer at all.

When I run a recruiting team, I also build the funnel per recruiter: candidates added, candidates contacted, HR interviews, hires — four stages and three conversions, each recruiter next to the others. Three pictures come up again and again:

1. Few candidates added. Either time is lost on people who never fitted (you will see a low conversion from presenting the vacancy to the next steps — teach how to qualify candidates), or time is managed badly, or the person simply does fewer actions than the others.
2. Many candidates added, few move on. Check how deeply the recruiter works with each candidate. The usual causes are the same message sent to everyone and a boring presentation of the vacancy on the call.
3. Many candidates, many presentations, few hires. Candidates like the vacancy but delay: there is no sense of urgency, or the recruiter only touches surface needs and never gets to the candidate’s real ones.

And one habit that turns the funnel from a picture into a diagnosis: record the reason every time a candidate drops out. Without reasons you know where people leave, but not what to change.

Top of the funnel. Plan the loading by week and by source, and check that the shares of the sources add up to 100% — this shows at once that the whole vacancy hangs on one channel. If the top stays thin, estimate the market before pushing the recruiter. The example from the course: we process 100 candidates a month, the market is about 5,000 people, and maybe only one to three of them will agree to our conditions. At this pace it takes about 25 months to reach our one, assuming that person sits somewhere in the middle of the list. With numbers like these the fix is in the requirements or the pay, not in the recruiter’s effort.

Middle of the funnel. Count the same weekly loading for the phone screening, the interview and the final interview. A low conversion here is usually about qualification (the wrong people were invited) or about the conversation (the right people were not interested).

Bottom of the funnel. Before an offer goes out, I use a ten-point checklist, one point for each “yes”:

1. Salary expectations match what the company can pay.
2. We know the possible start date and it suits us.
3. We discussed a counteroffer, and the candidate will not accept it.
4. We know the candidate’s motivation and can meet it.
5. We know which other offers are being considered, and agreed that the candidate contacts us before accepting any.
6. Our value proposition looks better than the current employer’s.
7. There are no red flags left unclear.
8. The candidate actively cooperates and shows interest in the vacancy and the company.
9. The candidate is ready to decide and does not plan to keep searching.
10. The hiring manager has shown interest in this person as a member of the team.

Every zero on this list is a possible reason for a rejected offer. Find it before the offer, not after.

The simplest report I have used, and the most useful: one tab per vacancy in a spreadsheet, two tables on it.

• Activity — for the week and for all time: responses from candidates, candidates found by active search, phone screenings, HR interviews, tests, interviews with the hiring manager, final interviews, offers made, offers accepted.
• Candidates — name, status, comment.

A hiring manager who is happy with the candidates rarely asks for reports. The one thing in a report that is really useful to them is the funnel: it shows that enough candidates were looked at, so the one they like can be hired with a clear conscience. At iDeals we had a rule for this — a decision on the hire was made only when we could say we were considering three candidates, not fewer.

For the wider set of metrics around the funnel, see the recruitment dashboard guide.

A funnel tells you where candidates are lost. What they felt at that stage is a separate question — I cover it in the guide to candidate experience design. The full path — funnel analysis, sales techniques for every stage, texts, communication with candidates and difficult vacancies — is in my Candidate Experience Course: Recruiting Funnel Training.