FAQs: Quality management

How do I choose appropriate strictness values when entering a topic phrase?

To determine the optimal strictness for a phrase, start with a default setting, then evaluate the captured events and adjust the strictness accordingly. Experimentation is key to finding the right balance between precision and recall.

Longer phrases with more meaningful words often require less strict matching. For example, a three-word phrase might match effectively with medium strictness (two out of three words), while a five-word phrase might only need three matches (medium-low strictness). The ideal strictness depends on the specific phrase and its intended use.

Below are some examples where adjusting strictness can either help or hinder your results, highlighting the importance of tweaking strictness for the best precision/recall tradeoff.

For more information, see Work with a topic, and Work with a phrase.

Examples where raising the strictness will avoid mistakes (increase precision)Click to expand

Defined program phrase	Detected phrase	Topic matched	Selected strictness	Explanation
Called the other day	Called there the other day	Contacted Previously	Medium	“There” provides context that make this irrelevant to the matched topic. Increasing the strictness would deem this occurrence as ‘not a close enough match’ for Contacted Previously.
I called back	I may call back	Contacted Previously	MedLow	“May” provides context that make this irrelevant to the matched topic. Increasing the strictness would deem this occurrence as ‘not a close enough match’ for Contacted Previously.
You’d like to cancel	Cause it wouldn’t be like cancelling	Cancel Mention	MedHigh	The context of “cancel” is different, therefore making this irrelevant to the matched topic. Increasing the strictness would deem this occurrence as ‘not a close enough match’ for Cancel Mention.

Examples where raising the strictiness will miss good occurrences (decrease recall)Click to expand

Defined program phrase	Detected phrase	Topic matched	Selected strictness	Explanation
I called a while ago	Called a little while ago	Contacted Previously	Med-Low	Contacted Previously was correctly matched, despite the extra “little” in the detected phrase. Increasing the strictness would result in classifying this occurrence as “not a close enough match” for Contacted Previously.
I spoke to another person	Person I spoke to	Contacted Previously	Med-Low	Contacted Previously was correctly matched, despite the missing “another” in the detected phrase. Increasing the strictness would result in classifying this occurrence as “not a close enough match” for Contacted Previously.
Want to cancel	Want to just cancel	Cancel Mention	Med-Low	Cancel Mention was correctly matched, despite the extra “just” in the detected phrase. Increasing the strictness would result in classifying this occurrence as “not a close enough match” for Cancel Mention.
Cancel the order	Cancel the money order	Cancel Mention	Med-Low	Cancel Mention was correctly matched, despite the extra “money” in the detected phrase. Increasing the strictness would result in classifying this occurrence as “not a close enough match” for Cancel Mention.

What is the retention period for quality management evaluation data?

Currently, Genesys Cloud keeps all quality management evaluation data for an indefinite period.

Topic spotting – What is the best way to organize my topics ?

The best way to organize topics is to tag them.

Tags provide a versatile framework for categorizing and classifying topics across a wide range of contexts. Tags create contextual connections between seemingly unrelated topics. By tagging topics, you establish an intricate web of associations, enabling users to discover related information. For example, tags can be used to find topics in Topic Definitions.

Which evaluation policy should I use?

The selection of one of the following evaluation policies depends on the individual use case and process your organization follows.

Assign Evaluations by Interactions – Automatically assigns an evaluation for every matching interaction. You can specify a form and one or multiple evaluators. If multiple evaluators are specified, they will each be assigned the same set of evaluations to complete.
- Use case(s):
  - A subset of agents should have all their interactions evaluated (for example: new hires).
  - Compliance monitoring that requires that every interaction with potential issues should be evaluated.
Assign Evaluations by Evaluators – Limits how many evaluations Genesys Cloud assigns to the specified agents, per the user defined interaction matching criteria.
- Use case(s):
  - When you want to ensure a specific number of evaluations to be assigned per evaluator in a given time interval. For example, you can configure this policy to generate and assign 10 evaluations per selected evaluator for a specific type of interaction, regardless if the number of interactions that match the criteria is greater than 10.
Assign Evaluations by Agents – Automatically assigns the specified number of evaluations per period, for each agent listed in the Matching Criteria.
- Use case(s):
  - When you want to ensure that an even distribution and a specific number of evaluations per assigned agents are generated. For example, if your weekly quota per agent is 2 evaluations, you can setup this policy type to assign 2 evaluations every week, to use the specific form of your choosing and, to assign them to a specific evaluator or set of evaluators.

Note:

When a policyhas an action to Assign Evaluations by Evaluator, Assign Evaluations by Interactions, or Assign Calibration Evaluations, the policy selects the first agent connected to the interaction as the agent to be evaluated.
When a policy creates an evaluation for Assign evaluations by agents , the policy selects the last agent that participated in the interaction. The policy will not select a user who was a monitor or coach.

What is quality management ?

Quality management is a set of tools and processes to ensure that all interactions are handled effectively and efficiently. Quality management includes interaction recording, real-time monitoring, search tools to locate recorded interactions, interaction evaluation and reporting, and calibration features. For more information, see About quality management.

How do I search for interactions that contain a transcript?

Use the Content Search view to find interactions that contain a transcript. To view interactions with transcripts, click Performance > Conten t. Click Menu > Analytics > Analytics Workspace. In the Default section, search for Content Search and then click the view name to open it. You can use the Transcript Content filter to search for specific words that are included or not included in a transcript.

How can managers see their employee’s quality evaluations?

Anyone with the appropriate permissions can see evaluations, so it depends on what permissions your managers have. To see evaluations, you need the Quality > Evaluation > View permission.

If your managers have the Genesys Cloud supervisor role with default permissions, they can view evaluations through the Agents Evaluation Summary or Agents Evaluation Detail views.

To see those views, users need the following permissions:

UI > Supervisor Agent Details – View
Directory > User – View
Quality > Evaluation – View

By default, the Genesys Cloud Supervisor role includes the first two permissions. You can either add the last permission to the Genesys Cloud Supervisor role or create a new role with that permission and add all of your managers to that role. For more information, see Quality management roles and permissions.

What happens if a metered evaluation policy matches more interactions than it is set to evaluate?

Genesys Cloud does not assign evaluations for the extra interactions. For example, imagine that you create a policy that matches inbound calls and that also assigns 50 metered evaluations per day. If you receive 300 inbound calls in a day, then only 50 of them are assigned for evaluation. The remaining 250 are no longer eligible for evaluation, unless they match a different policy. The next day, the count starts over with new interactions received that day. The non-evaluated interactions from the previous day do not roll over.