Stay up-to-date with every new feature in Qualtrics XM

LEARN MORE

Ready to try Qualtrics?

GET STARTED

Getting Started with Intelligent Scoring

Suite

Customer Experience

Product

Qualtrics

What's on this page

About Intelligent Scoring

Intelligent scoring assesses your XM Discover interactions with categorization and flexible rules-based scoring. Intelligent scoring scores subjective interactions on quantitative scales, simplifying the complex and allowing your workforce to focus their efforts on the most important insights.

Intelligent scoring can be useful for:

Performing quality management of customer-facing agents, helping assess and coach agents on soft skills like professionalism, empathy, and knowledgeability.
Qtip: Intelligent scoring also helps remove some of the bias that customer-given scores like CSAT and NPS introduce. For example, a customer may be upset about a return or product issue, and negatively rate an agent out of pure frustration. In contrast, intelligent scoring only takes the agent’s performance on soft skills into consideration when giving them a rating.
Tracking public reaction to current events on social media, empowering your company to decide how best to accommodate customers and make evidence-based decisions in times of uncertainty.
Determining legal risks.
Determining sales efficacy.
And more customer experience use cases!

Setting Up Intelligent Scoring

In order to set up intelligent scoring, you’ll need to complete a number of steps in both Designer and Studio.

Select a scoring model: In Designer, select a category model to use for scoring.

Qtip: If you want to score point-forward data as it’s ingested in XM Discover, select Include in Incremental Data Loads.

Create a rubric: Rubrics are how you set scoring expectations for documents. In Studio, define your criteria, assign them weights. You can even assign items that automatically fail subjects, such as unprofessional conduct.

Enable a rubric: By enabling a rubric, you make it available for scoring and reporting.

Rescore historical data (optional): Apply intelligent scoring to historical data or update previously scored documents to reflect changes in rubrics or topics.

Create a scorecard report: Use intelligent scoring in reports.

Navigating to Intelligent Scoring

Log into Studio.

Open the global navigation menu.

Intelligent Scoring from the global navigation menu

Click Intelligent Scoring.

Select a project from the dropdown menu.

The list of intelligent scoring rubrics for the project you select will be displayed in the table. From this table, you can create, edit, and manage rubrics.

Areas of the intelligent scoring navigation table

Search for a rubric in the project. A project must be selected before you can search for rubrics.

Edit which columns appear in the table.

Sort the table of rubrics. Alternatively, you can click on a column header to sort the table.

Create a new rubric. For more information, see Creating Rubrics.

Edit, disable, rename, or delete a rubric.

Enable and disable rubrics.

FAQs

What is a category model? What is a topic?

A topic is a specific theme mentioned in open-ended feedback that XM Discover captures using category models. A category model is a rules-based hierarchical taxonomy used to organize sentences into topics.

Since category models are how XM Discover analyzes topics, you will see “category model” and “topic” used interchangeably throughout the platform.

What is the difference between a rubric and a scorecard?

A rubric defines the expectations, while a scorecard illustrates the results.

One way of thinking about it is that a rubric is the input, and a scorecard is the output.

Can I create more than one rubric?

Yes! You can use up to 10 models per project to set up the rubrics. You can use many rubrics per model. Any given topic (including its siblings and children) can only be used in one rubric.

Being able to use multiple models to define rubrics is particularly important because the root node rule of a given model is the first form of conditional logic. If a document does not meet the root node rule, it will not be scored. This is so it will not generate false positives for alerting and case creation.

Are there any recommendations around assigning weights?

If your organization has an established quality management process, we recommend reusing your rules in XM Discover. If not, we recommend using drivers.

Here’s how you can use drivers for quality assessment:

Use the scoring model as an input for drivers and set the outcome to be a high CSAT score or customer experience score.
See which topics and behaviors have the highest impact.
When creating a rubric, select the driver from the Suggest Ranks drop-down to display topics' impact rank next to their weights.
Assign larger weights to topics that are more likely to drive to or away from a desired outcome. Another approach is to start with equal weighting, then adjust it based on your needs.

If the rules used in a rubric are present multiple times in a document, will that result in multiple scores?

No. Scoring is done at the document level. If a rule is hit at any point, it is scored then, only once.

What's the behavior of the intelligent score attribute during the rescoring process? Does it show as "null" for all documents, or does the old score stay until the new score replaces it?

The documents' scores only update once they have received their new scores. There is no "null" period in-between, so they remain visible in reporting.

An exception to that would be if there was some update to the scoring logic that excluded some documents that previously had scores. In this scenario, those documents would be updated to "null."

If I classify a rubric model, does that have the same effect as rescoring a rubric? How does it impact historical document scoring?

Performing a classification and rescoring a rubric have the same effect. The only difference between a rescore and classification is that with rescoring, you can target a subset of data (using a filter) without affecting the scores of data that does not match the filter. This can serve the purpose of performance optimization and rudimentary scorecard versioning.

Was this helpful?

The feedback you submit here is used only to help improve this page.

That's great! Thank you for your feedback!

Thank you for your feedback!

Getting Started with Intelligent Scoring

About Intelligent Scoring

Setting Up Intelligent Scoring

Navigating to Intelligent Scoring

FAQs

Get the latest news on XM events, research, and product launches