What are the possible predictions for a job's target start date?
|Not enough candidates||We're not able to make a prediction until a job has at least 1 candidate. Add a candidate to your role to see a prediction!|
|No status||You haven't set a target start date yet. Click Set Target Start Date to begin.|
|On track||Your predicted start date falls before or the week of your target start date. You're on-track to make a hire by your target!|
|Heads up||Your predicted start date falls after the week of your target start date. You're not on track to make a hire by your target and should consider adding more quality candidates to your pipeline.|
|Not confident||We only display a prediction on the job dashboard if it meets certain accuracy standards. Check back later when we've had more time to crunch the numbers.|
|Error||One of our internal systems powering Greenhouse Predicts is not working properly. We have monitoring in place to let our engineers know when there is a problem, and we're working on getting Greenhouse Predicts back up and running!|
Who can see predictions on the job dashboard?
Any Job Admin on the job will be able to see predictions on the job dashboard. However, only Job Admin with the permission 'Can edit job info' will be able to set target start dates.
My predicted start date range is always 4 weeks long. Will this range get shorter and more accurate over time?
We make predictions for the offer acceptance date because there is usually less variability for when someone will accept an offer than when they'll be able to start working. For example, a candidate might need to relocate, wait for a visa, or give notice to their current company. All of these factors make it harder to accurately predict a start date.
Therefore, we predict the week a candidate is likely to accept an offer. Then, we give you a start date range that is always a fixed timeframe of 4 weeks, starting 2 weeks after our predicted offer acceptance date. This range will not get shorter over time, but you can use it as a guide for when to expect a new hire to start.
How do I know which job opening the predictions are for?
We currently only make a prediction for one opening at a time on a job.
We can confidently make a prediction for one hire based on the job's pipeline, but our accuracy would significantly decrease if we tried to predict for two or more hires from the same pipeline. We'd be using the same pipeline to make predictions, but trying to determine multiple outcomes which would increase variability and lead to less-precise predictions.
If you have more than one opening on a job, you can still set multiple target start dates. We’ll make a prediction for the opening with the soonest target start date. Once you make a hire, we’ll start predicting for the next opening on that job right away!
How accurate is the prediction I'm seeing?
On average, 80% of the time, our predictions are within 3 weeks of true time-to-next hire.
Because Greenhouse Predicts was built using machine learning, predictions will continuously improve and become more fine-tuned to your company over time as we get to know your hiring habits.
Additionally, we're actively evaluating and testing new model inputs that will increase the accuracy of our predictions.
Why is my screen is showing a different prediction than my team member’s?
Because we use machine learning to make predictions, there is a chance that the equation can weigh factors slightly differently each time we run this calculation.
When you look at your job dashboard, and someone sitting next to you looks at the same job dashboard on their computer, we're actually running two different calculations.
This means that our predictions might bounce around a little bit from week to week, especially if the prediction is right on the edge of the end of one week and the beginning of another week. This can cause some variability in the prediction you might see on your screen.
Why does the prediction jump forward or back a week just by changing the target start date?
Because we use machine learning to make predictions, there is a chance that the equation can weigh factors slightly differently each time we run this calculation (i.e. anytime you view a prediction on the job dashboard).
This means that our predictions might bounce around a little bit from week to week, especially if the prediction is right on the edge of the end of one week and the beginning of another week. This can cause some variability in the prediction you might see on your screen when the Greenhouse Predicts widget refreshes after setting a target start date.
Does setting a target start date impact my prediction?
No, we will still make predictions for a job even without a target start date. Setting a target start date will not impact the predictions, but you won't get an on track or heads up status until you set one.
Why can't I see historical predictions?
We do not store historical predictions over time in Greenhouse Recruiting.