Recruitment Funnel

What a recruitment funnel is

A recruitment funnel is a representation of the successive stages of hiring, from application to acceptance of the offer, showing how many candidates reached each stage, so the figure narrows from one stage to the next. It is called a funnel because the counts fall as the process advances: more people apply than are hired. What is useful in it is not the shape but the ratios between one stage and another, because they show where the counts fall and by how much.

The funnel answers one question for a recruiter: where do the candidates leave? That means at which stage the counts fall by more than expected, which is where to look. It does not say why they leave, because the reason is learned by examining those who left and not from the numbers.

The stages of a recruitment funnel and what is counted in each

• Application. Those who sent an application. What comes before, meaning those who opened the advertisement or began an application and did not send it, is measured by the application completion rate, and the application abandonment rate measures the other side of the same figure.
• Screening. Those who passed the review of the application or the first screening interview, and the prescreening interview is one of its tools.
• Interview. Those who were invited to one interview or more.
• Offer. Those to whom a job offer was made.
• Acceptance. Those who accepted the offer.

These stages are not fixed in every organisation. One organisation adds a test after screening and another drops an interview. What matters is that a clear entry rule is written for each stage, so a candidate is not counted in “interview” merely because a time was set, unless that is the rule and it is applied to everyone.

An example of a quarter’s funnel divided into two categories

Take assumed figures that illustrate the calculation: 800 people applied for the organisation’s jobs in a quarter. Of them 240 passed screening, 96 were interviewed, 30 received an offer and 24 accepted. The ratios between stages are:

• From application to screening: 240 ÷ 800 = 30%.
• From screening to interview: 96 ÷ 240 = 40%.
• From interview to offer: 30 ÷ 96 = 31.25%.
• From offer to acceptance: 24 ÷ 30 = 80%.
• From application to acceptance: 24 ÷ 800 = 3%.

The overall figure is not enough, because the organisation hires in more than one job category. If the quarter is divided into administrative and technical jobs, what was hidden appears:

• Administrative. 500 applied, 150 passed screening, 60 were interviewed, 18 received an offer and 15 accepted. That is 3% from application to acceptance, and an acceptance rate from the offer of 15 ÷ 18 ≈ 83.3%.
• Technical. 300 applied, 90 passed screening, 36 were interviewed, 12 received an offer and 9 accepted. That is 9 ÷ 300 = 3% as well, and an acceptance rate from the offer of 9 ÷ 12 = 75%.

The overall 3% (application to acceptance) and 80% (offer to acceptance) therefore combine two categories that are identical in the ratio from application to acceptance, although acceptance of the offer is somewhat lower in the technical category. With such a small number of offers, twelve and eighteen, the gap between 75% and 83.3% is too small to base a decision on, because one more candidate accepting in the technical category would lift 75% to 83.3%, the administrative figure. The sound reading is to look at the funnel by category and by channel and to give the counts from which each ratio was derived, not the ratio alone.

The example shows two things that the overall figure does not: the first two ratios are identical in the two categories, and the difference appeared in the last two stages. The reason for it in the technical category is not told by the number, and it is learned by asking those who did not accept the offer.

A cohort funnel and a period funnel

A funnel can be counted in two ways:

• A cohort funnel. The group of those who applied in a given month is followed until their cases end, and the count says how many of them reached each stage. Its ratios are complete only after the cases of the group have ended.
• A period funnel. It counts what happened in the month in applications, screening, interviews, offers and acceptances, whenever each person applied. It shows sooner, but it mixes those who applied in two months.

Take a month in which 100 applications were recorded and 150 candidates reached an interview, because candidates who applied in the previous month were interviewed in this one. The ratio between application and interview becomes 150%, a figure with no meaning because the numerator and the denominator come from two different groups. Each funnel is therefore stated as a cohort funnel or a period funnel, and otherwise its reader may think that what is seen applies to a single group of people.

How a recruitment funnel differs from a candidate pipeline

The candidate pipeline is defined as the set of candidates moving towards one specific vacancy, distributed across defined stages, and it begins when the requisition is approved. The pipeline belongs to one vacancy and is managed while the vacancy is open, so the question in it is about a particular candidate and where that candidate stands now, and that term also covers the pass rate between each two consecutive stages in that vacancy.

The funnel looks at the same ratios, gathered from more than one job across a period, after the cases have accumulated, so the question in it is about counts and ratios and not about people. The two complement each other: the cases of closed pipelines can be collected into a funnel, and the funnel shows what a single vacancy does not show, which is what recurs across vacancies. The tool that records the stages for both is described under applicant tracking system.

How a recruitment funnel is built from system data

A funnel rests on the date on which each candidate entered each stage and on the reason for leaving, if the candidate left. Two conventions are possible: counting a candidate only at the highest stage reached, or counting the candidate in every stage reached or passed. Suppose a candidate is referred by an employee and invited straight to an interview with no screening. Under the first convention the candidate appears in the interview and does not appear in screening, and the ratio of interview to screening can exceed 100%. Under the second the candidate enters screening as well. Only a declared rule for skipped stages fixes the ratio, so the rule is stated with the figure.

It is also necessary to keep the reason for leaving at each stage, for example that the candidate did not meet a condition, withdrew or declined the offer. Without the reasons the funnel is only counts, and nothing can be put right from it. A tracking tool that records the transitions, such as an applicant tracking system, gathers this data.

How a recruitment funnel is used in a decision

• Choosing where to put things right. The ratio at each stage is compared with the organisation’s own earlier values, so the stage whose ratio has changed by the greatest amount is examined, not the one with the lowest ratio.
• Comparing channels. Funnel ratios are calculated for each channel, which shows which of them gives candidates who reach the offer at a higher rate.
• Estimating the number needed. If the ratios hold steady, the number of applicants needed for a given number of hires can be calculated. In the example above the ratio from application to acceptance is 3%, so hiring six needs 6 ÷ 0.03 = 200 applicants. This is an estimate resting on the ratios staying as they were, a condition that may not hold, so the figure serves as a starting point for planning and not as a promise.

What a recruitment funnel does not show

• The reason for the narrowing. The screening stage can narrow because the advertisement attracts people to whom the requirement does not apply, or because the criterion is narrower than the job needs. The number does not tell them apart, and the recorded reasons for exclusion do.
• Whether a narrowing is a fault. A stage that removes more than half of those who enter it may be doing its job, so a low ratio is not enough to count as a problem.
• The quality of those who got through. A funnel in which acceptance from the offer reaches three quarters may have made its offers to unsuitable candidates. Quality of hire measures the result after months.
• The source of the candidates. The overall funnel merges channels. Source of hire distributes those hired across their channels, and funnel ratios are then calculated for each channel.

How a recruitment funnel relates to the selection ratio

The selection ratio, the number hired divided by the number who entered the selection stage, is examined alongside the funnel. Its result changes with the denominator chosen, which is the same decision made in defining the entry rule for each stage.

Mistakes in reading a recruitment funnel

• Adding together jobs that are not alike. One funnel for a job that draws hundreds of applications and a rare job that draws only dozens hides both, since each kind has its own ratios.
• Counting one applicant more than once. Someone who applies for three jobs may be counted as three applications, which inflates the denominator of application and lowers the first ratio although nothing has changed in the advertisement. The number of applications is kept apart from the number of people.
• Improving a ratio by loosening the criterion. Raising the screening ratio by passing those who do not meet the condition makes the first stage look better and loads the next stage with people who should not have reached it. The ratio of each stage is therefore read together with the stage after it.
• Ignoring time. The funnel does not show how long a candidate spent between one stage and another, and an organisation that keeps its ratios and lengthens the wait loses candidates who show up in the ratios only when they leave.

What counts as a normal ratio between funnel stages

The ratio between two stages follows the job, its market and the entry rule for each stage, so no particular ratio is a standard for all organisations. If someone says that the right screening ratio is a certain figure, the questions are how screening was defined and from which group the ratio was taken. In the sources we reviewed, we found no published reference for these ratios in the Saudi market.

Each ratio should therefore be written with the entry rule of the stage it starts from and of the stage it ends at. If a ratio changes over a period, the first question is whether the entry rule changed, and then whether the advertisement, the criterion or the party sending the candidates changed.

This is an explanation of the concept, not legal advice.