Time to Hire

What time to hire is

Time to hire is the number of days that a particular candidate takes from entering the hiring process until accepting the job offer. It is therefore a measure of the candidate’s journey and not of the age of the vacancy. Its start depends on the definition adopted: the candidate submitting an application, or the first contact with the candidate where the contact began with the organisation. It ends when the candidate accepts the offer, not when the candidate starts work.

It shows a recruiter how long a candidate waits between steps in the process, where the delay falls, and whether the journey differs from one channel to another. Because its start and end can each be placed in more than one spot, any figure needs its two ends stated before it is compared with anything.

How time to hire differs from time to fill

Time to hire measures the journey of one candidate, while time to fill measures the age of a vacancy. Time to fill is a measure of the vacancy: it starts when a requisition is opened and is taken to end when the new hire starts work. Time to hire starts later, with the candidate. For the same vacancy, time to hire can be shorter than time to fill, because the days before the candidate arrives, such as approving the requisition and publishing the advertisement, do not enter it.

• Unit. The candidate for time to hire, and the vacancy for time to fill.
• Start. The candidate entering the process, not the approval of the requisition.
• End. Acceptance of the offer. Time to fill is taken to end at the start of work.
• What it tells. How long this candidate waited, not how long the department stayed without an employee.

The name causes a second confusion. The explanation of cost per hire uses a measure called “time to hire” that counts the days from the vacancy opening to the acceptance of the offer, which is a different start from the start of time to hire here. Two documents can give the same name to two measures, and their starts tell them apart. For that reason each figure is written with its two ends.

The start decides the result: an example with candidates from two channels

The importance of fixing the start shows when candidates enter the process by two different doors. The figures here are assumed, to show the structure. Six candidates were hired. Three applied on their own, and three were first contacted by the recruiter and then asked to submit a formal application.

• Those who applied on their own. Between application and acceptance of the offer the times were 18, 24 and 30 days, an average of 72 ÷ 3 = 24 days.
• Those who were contacted. Between first contact and acceptance the times were 35, 41 and 47 days. They submitted their formal applications 28, 30 and 31 days after first contact, so between application and acceptance the times were 7, 11 and 16 days, an average of 34 ÷ 3 ≈ 11.3 days.

If the count starts from application, the average of the two groups together is (72 + 34) ÷ 6 ≈ 17.7 days, and the contacted candidates look faster than those who applied on their own, 11.3 days against 24. If it starts from first contact, the average is (72 + 123) ÷ 6 = 32.5 days and the order reverses: 41 days for the contacted candidates against 24. The candidates and the process are the same, and the result changed because the start changed. An organisation that compares two channels whose counts begin at different points is judging the channels by their start and not by their performance.

The rule that follows is to adopt one start for all candidates and to record both the date of first contact and the date of the formal application in the system, so both figures can be calculated and shown. Talent sourcing is the reaching of candidates before they apply, and it can produce this difference in the point of start.

How time to hire is calculated for a group of candidates

It is calculated for each candidate by subtracting the start date from the date the offer was accepted, and the results are then combined at group level. Decisions in the calculation change the figure, among them:

• The unit of days. Calendar days or working days. The first is simpler, and the second is closer to what the candidate actually waits when holidays intervene. Whichever is chosen is used in every period.
• The statistic. The mean or the median. A single long case raises the mean and does not raise the median, so the median is shown with the mean and the longest case, as the explanation of time to fill does for its own measure.
• The group. A case is attributed to the month in which the offer was accepted, not the month in which the candidate applied. If cases are attributed to the month of application, the latest month of application looks faster than it is, because those who have not yet accepted have not entered it.

The last decision is the least visible. Suppose ten candidates applied in one month, six of them accepted in the same month, and four remained in the process. Anyone who calculates time to hire at the end of the month on the six leaves out four whose wait will grow longer, so the figure looks shorter than it will be when the month is closed on all ten. A month’s figure should therefore be compared with the previous month’s only once the month’s cohort has closed.

Who is left out of the time to hire calculation

Time to hire is calculated only for those who accepted the offer. Candidates who withdrew in the middle of the process, who declined the offer or who were excluded do not enter the figure, even if they are the ones who lengthened the process. The figure can therefore look better as more candidates withdraw after a long wait.

If twelve candidates reach the offer stage, and eight accept and four decline, time to hire is calculated on the eight alone. If the four waited longer than the eight, the figure would not show it. Acceptance by eight of twelve, which is two thirds, helps in reading the figure and does not replace it, so the number who reached the offer and the number who accepted are shown beside time to hire. The offer at which the journey ends is described under job offer.

Where the delay falls within time to hire

The overall figure does not show where the delay is. What shows it is dividing the figure by the stages of the candidate’s journey, according to the dates of each stage, and taking the median of each stage and not its mean, because one long case distorts a mean. A tool that records the date of each move helps with this, and it is what an applicant tracking system offers, while the candidate pipeline describes the stages into which the journey is divided.

The stage with the longest delay may not be the stage with the greatest amount of work. The delay may be a wait between two steps, such as waiting for an interview slot or for a signature on a decision. In those days nobody is working on the candidate; they appear in no procedure and show up only in the gap between the dates of two consecutive stages. The remedy may be an arrangement of who decides and when, not a faster screening.

What is recorded for each candidate

To allow any definition to be calculated later, the basic dates are recorded for each candidate: first contact where there was one, the application, entry into each stage, the making of the offer, and its acceptance or refusal. If the dates are missing, only what was recorded can be calculated, and a date that has passed cannot be recovered. An organisation that records only the acceptance date knows where the candidate ended up and does not know where the candidate spent the wait. These dates can be held in an applicant tracking system where one is used, and the check is that each stage is recorded with its date and not only with its status.

After acceptance: notice and starting work

Time to hire ends at acceptance of the offer, and what follows is outside it. Between acceptance and starting work there is a period that may be the longest in the journey, because the candidate may be at another job and owe the employer a period of notice, an obligation described under notice period. The time to hire measure decides nothing about that period, so it is best used alongside time to fill, which runs to the start of work and suits better whoever wants to know how long the position stayed without an occupant.

Why the time to hire figure moves

If time to hire falls between two periods, different causes may be producing it, and only one of them is a real improvement:

• The process sped up. The gaps between stages shortened. That is the improvement, and it is verified by comparing the stage times and not the overall figure.
• The people entering the calculation changed. Fewer and faster candidates accepted offers, and those whose wait would have been long withdrew. The figure falls while the organisation has lost people who might have joined.
• The start changed. The point of counting was redefined, as in the example above, so the figure moved without the process moving.

Each fall should therefore be checked. Have the stage times come down, has the number who accepted held steady, and has the start stayed as it was? If the answer to any is “no”, the figure is not evidence of improvement.

How time to hire relates to neighbouring measures

• Candidate experience. The length of the wait affects what the candidate feels, and candidate experience covers that side. The figure is a starting point for the question of experience and not a substitute for it.
• Quality of hire. It may be thought that the shortest period is the best, but quality of hire measures the result after months, and we have no measurement on which to say that shortening the time raises or lowers quality.
• Source of hire. Time to hire can be divided by channel to show which channel is faster, and source of hire shows how a hire is attributed to its channel. Channels are compared on their time only if the start of the count is the same in all of them, as in the example above.

What period counts as acceptable for time to hire

The period varies with the occupation and its state in the market, and with the start adopted, whether application or first contact, and the end at acceptance of the offer. A figure for a “good time” that is not given with its two ends is not suitable for comparison. In the sources we reviewed, we found no published reference for the durations of this measure in the Saudi market, and no text that requires an establishment to measure it.

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