Get Predictions About the Time Required for a First Hire

When the Time to Hire feature is enabled within your organization, you can get predictions about the time it will take to make a first hire for a job requisition. The Time to Hire feature uses Artificial Intelligence (AI) and machine-learning algorithms to estimate the time for a first hire, based on previous similar job requisitions.

The Estimated Time to Hire section is available when you create a job requisition and edit a draft job requisition, so that you can see the estimated time to hire before finalizing the details of the requisition.

When you start creating a job requisition, only the estimated time to hire is displayed. When the requisition reaches the Open phase, another number is displayed to indicate the current number of days for which the requisition has been open, letting you compare the current time with the estimated time to hire. After a first hire is made on the requisition, the current number of open days is no longer displayed. It's replaced by the number of days it took for a first hire to be made.

You can use the feature to model scenarios where the location or education required for that position is changed. For example, you’ve posted a role for an Engineer in San Francisco and you view a prediction of 42 days to hire. You can then model a scenario where you change the location to California to see how the days to hire prediction is impacted. You can change the value of these 3 fields displayed in the Estimated Time to Hire section so you can see the impact it has on the time to hire prediction.

• Requisition Title

• Education Level

• Locations

When you change these values, it has no impact on the actual requisition values. You need to change the corresponding values in the requisition details if you want to keep them.

How Time to Hire Works

The Time to Hire model trains on all job requisitions that had a hire in the last 365 days, and each job requisition can influence the prediction outcome. It not only considers the length of time that a job requisition was open, but also considers factors such as the job location and title. Therefore, the time to hire predictions are more than just an average of the time it took to close historical requisitions.

The Time to Hire feature uses a supervised machine learning model known as Random Forest. This classification algorithm works by generating and comparing all the job requisition input data with all possible variations of that same input data. After these computations are made for all possible variations, an average is taken across the entire set to generate a prediction.

After every job requisition is closed, the model uses the days required to close the requisition to improve the quality of future recommendations. The Time to Hire model trains on this and other new data once a week. So, if you’re hiring within a short period of time, you can’t expect an immediate change in the predication.