> ## Documentation Index
> Fetch the complete documentation index at: https://help.journeyfront.com/llms.txt
> Use this file to discover all available pages before exploring further.

> ## Agent Instructions
> This is the Journeyfront Help Center for hiring teams. Journeyfront is an applicant tracking system (ATS). Use the Help Center's terms: job req, Candidate Profile (the person), Candidate Application (one person on one job req), hiring class. Link to the article you used in your answer.

# Predictive models

> Use models built from your own employee outcomes to predict which candidates will perform and stay, and include the prediction in their score.

> [!NOTE]
>
> **Who can do this:** Users with the *Manage Job Req Settings* permission add predictive models to a job req and set how much they count.
>
> **Availability:** Journeyfront builds predictive models for your job profiles from your employee outcome data. To get one for a role, ask your Customer Success Manager.

A predictive model estimates how likely a candidate is to succeed in a role, for example how well they'll perform or how likely they are to stay. Journeyfront builds each model from your own employees' outcomes, such as performance and turnover, in that [job profile](https://help.journeyfront.com/employees/job-profiles.md). On a job req, you choose which models to use, and every candidate gets a predictive score that can count toward their overall score.

## How predictive models fit together

**1. Collect outcome data**

Your employees' outcomes in a role, like performance and turnover, are the data a model learns from. The [behavioral assessment](https://help.journeyfront.com/assessments/behavioral-assessment.md) is made to go with predictive models.

**2. Journeyfront builds the model**

Each model belongs to a job profile and predicts one outcome. You'll see it under **What it Predicts**, for example **Performance**.

**3. You add it to a job req**

A job req can use the models built for the job profile it's linked to (**Linked to Job Profile**, under **Setup**, then **Details**).

**4. You include it in the score**

Turn on **Use Predictive Model** and set the weights in the job req's **Setup**, then **Configurations**.

> [!TIP]
>
> No model for a role yet? You can still use the behavioral assessment on its own: set the ideal value for each trait you care about, and start collecting the data a model can be built from later. See [Behavioral assessment](https://help.journeyfront.com/assessments/behavioral-assessment.md).

## Turn on predictive models for a job req

### 1. Add the models

**1. Open Predictive Models**

Open the job req and go to **Setup**, then **Predictive Models**. The table lists the models the job req already uses, with **Date Created**, **What it Predicts** and **Weight**.

**2. Add a model**

Click **Add Model**. The **Select Predictive Model** window lists the models available for this job req's job profile. Click **Add** next to each one you want, then click **Save**.

**3. Wait for scores to update**

Journeyfront recalculates the scores of the job req's candidates. On a busy job req this can take a few minutes.

![The Predictive Models page under a job req's Setup tab, with the Edit Weights button, the Add Model button highlighted, and one model, Customer Care Agent Performance, created 03/12/2026, that predicts Performance](https://mintcdn.com/journeyfront/Ki8OScNKQAoeTHdj/images/scoring/predictive-models/gen-predictive-models-page.png?fit=max\&auto=format\&n=Ki8OScNKQAoeTHdj\&q=85\&s=9f9921490f7a065bb34cb8488cc31b2d)

If the window shows **No Data**, there's no model for this job req's job profile yet. Check the job req's **Linked to Job Profile** setting, then ask your Customer Success Manager.

If the job req uses more than one model, click **Edit Weights** to decide how much each model counts compared with the others, and click **Save**. A higher weight counts more.

To stop using a model, click the trash icon on its row, then **Confirm**.

### 2. Include the prediction in the overall score

**1. Open Configurations**

In the same job req, go to **Setup**, then **Configurations**, and find the **Scoring & Reporting** card.

**2. Turn on Use Predictive Model**

Switch on **Use Predictive Model**. **Scoring Weights** appear.

**3. Set the weights**

Set **Candidate Predictive Score weight** and **Custom Attributes weight** to decide how much the prediction counts compared with the job req's own attributes. Then click **Save** at the top of the page.

See [Job settings](https://help.journeyfront.com/job-reqs/job-settings.md) for the rest of the page.

## How the prediction feeds the overall score

Journeyfront rolls scores up as weighted averages: questions and tests into attributes, attributes into attribute groups, and groups into one score. The weights you set decide how much each piece counts.

| | What makes up the overall score |
| - | - |
| **Use Predictive Model off** | The roll-up of your attribute groups. |
| **Use Predictive Model on** | Two roll-ups: your attribute groups and your predictive models. **Scoring Weights** decide how much each one counts. |

Adding or removing a model, changing a model's weight, or changing **Scoring Weights** recalculates the candidates' scores. To set up the attribute side, see [Attributes](https://help.journeyfront.com/scoring/attributes.md).

## See the prediction on a Candidate Application

When **Use Predictive Model** is on and the job req has at least one model, the **Score** tab of each Candidate Application starts with **Candidate Predictive Scores**:

* The first row is the candidate's combined predictive score.
* Below it, each model has its own row and score.
* Expand a model to see the traits it looks at, with the candidate's value next to the ideal value for each one.

The candidate's attribute scores follow underneath, starting with **Candidate Attribute Scores**. See [Reading a candidate's scores](https://help.journeyfront.com/scoring/reading-scores.md#the-score-tab).

![The top of a Candidate Application's Score tab: Candidate Predictive Scores with a combined score of 3.8, the Performance model expanded to show each trait's candidate value and ideal value on a slider with a score, and the Candidate Attribute Scores row below](https://mintcdn.com/journeyfront/Ki8OScNKQAoeTHdj/images/scoring/predictive-models/gen-predictive-scores.png?fit=max\&auto=format\&n=Ki8OScNKQAoeTHdj\&q=85\&s=91441f560a4e847295337f30992425a5)

You can also put predictive scores to work elsewhere:

* **The Candidates page:** add **Predictive Model Scores** with **Edit Columns**, or filter by them. See [The Candidates page](https://help.journeyfront.com/candidates/candidates-page.md).
* **Automation rules:** use the **Predictive model score is** condition to act on a model's score. See [Automation rules](https://help.journeyfront.com/workflows/automation-rules.md).

> [!NOTE]
>
> A predictive score is one input. Use it alongside interviews, assessments and your team's judgment; people make the hiring decision.

## FAQ

> [accordion] Why don't I see Candidate Predictive Scores on the Score tab?
>
> Two things need to be in place: **Use Predictive Model** is on under **Setup**, then **Configurations**, and at least one model is added under **Setup**, then **Predictive Models**.

> [accordion] Can we build our own predictive model?
>
> Not in the app. Journeyfront builds and maintains predictive models from your outcome data. Talk to your Customer Success Manager about a model for a role.

> [accordion] Why is the Select Predictive Model window empty?
>
> There's no model for the job profile this job req is linked to. Check **Linked to Job Profile** under **Setup**, then **Details**, and ask your Customer Success Manager if you expected a model.

> [accordion] Do we need predictive models to use the behavioral assessment?
>
> No. The behavioral assessment is included in every account. Without a model, set the ideal value for each trait on the job req, and your results become data a model can be built from later. See [Behavioral assessment](https://help.journeyfront.com/assessments/behavioral-assessment.md).

> [accordion] Does adding or removing a model change existing scores?
>
> Yes. Adding or removing a model, or changing any of the weights, recalculates the scores of the job req's candidates. It can take a few minutes.

> [accordion] Can candidates see their predictive score?
>
> No. Scores are only visible to your team.

## Related

> [card] [Behavioral assessment](https://help.journeyfront.com/assessments/behavioral-assessment.md)
>
> The assessment made to go with predictive models.

> [card] [Job profiles](https://help.journeyfront.com/employees/job-profiles.md)
>
> The roles your predictive models are built for.

> [card] [How scoring works](https://help.journeyfront.com/scoring/how-scoring-works.md)
>
> Raw scores, the Job Adaptive Score and how scores roll up.

> [card] [Reading a candidate's scores](https://help.journeyfront.com/scoring/reading-scores.md)
>
> Where every score shows on a Candidate Application.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.