Tag topics manually

Manual topic tagging involves a human user mapping phrases in customer feedback comments to specific topics.

A topic tag is a link between a phrase in customer feedback and a Medallia Text Analytics topic. Most implementations include automatic tagging, which uses machine learning to associate phrases with one or more topics, and which creates tags automatically. You can, however, supplement that automatic processing by tagging phrases to topics manually. This is known as hybrid tagging.

Medallia strongly recommends updating topic rules instead of relying on manual topic tagging on individual comments, especially for larger volume programs, because manual tagging can only apply to a single phrase in a single comment. Instead, update topic rules so that changes apply to all comments that match the rule.

Manual tagging is most useful in programs with a low volume of feedback (such as Voice of the Employee and B2B), where the cost in time and labor to review every comment for appropriate tagging is not prohibitive. Additionally, while automatic tagging guarantees an 80-85% accuracy rate, some programs might require the higher accuracy provided by manual tagging.

You can create tags to specific phrases using existing topics, but you cannot create new topics through manual tagging.

Before you begin

Manual tagging works for fields in tag pools with existing topics. While you can use manual tagging to control which existing topics are associated with phrases and comments, you cannot create new tags.

For more information about configuring tag pools, see Creating a tag pool.

Additionally, assign the following reporting permissions to roles that need to review and modify tags in customer responses. For more information about managing role permissions, see Roles.

  • View Comment Topics — Enables the role to see the topics associated with each comment as a whole, and with each individual phrase.
  • Edit Comment Topics — Enables the role to remove or change tags, either by removing a topic from an entire comment or phrase, or by changing the topic associated with an individual phrase.

Reviewing and changing topic tags

Text Analytics displays topic tags directly in Responses Feed modules and Responses Form reports. For example, the following image shows customer feedback for two comment fields in a Responses Feed module:

Customer comments in a Responses Feed module

Click the pencil icon to open that feedback for manual tagging. The tags under the comment fields (Staff - Overall Satisfaction, Staff - Helpfulness, and Brand - Competitor Mentions) apply to one or more of the individual phrases in their respective comments. Click the X next to a topic to remove the tagging between that topic and all of the phrases in the comment.

Topics that apply to one or more phrases in comments

To see the topics associated with an individual phrase, click the tag icon to the left of a phrase. For example, the following image shows that the first phrase (highlighted in green), is tagged with the Staff - Helpfulness topic:

A topic associated with a phrase in a comment

To remove a topic association from a specific phrase, click the X next to that topic in the dropdown. To add a topic to a phrase, click the plus icon, then select a tag from the dropdown. Repeat this process to add one or more additional tags, and then click Apply. For example, the following image shows the topics available to be added to a phrase when the plus icon has been clicked:

Topics available to be added to a phrase

When you are done editing the topic tags for the phrases and comments in a record, click the pencil icon to return to view-only mode.

Experience Cloud uses the following A-fields to record information about topic tagging.

For more information about A-fields, see A-fields.

  • a_topics_for_[field] — A comma-separated list of the topics that match a specific field.
  • a_topics_original_for_[field] — A comma-separated list of the topics that match a specific field, before any manual changes.
  • a_topics_sentiments_for_[field] — A comma-separated list of the topics and sentiments that match a specific field.
  • a_topics_sentiments_latest_for_[field] — A comma-separated list of the most recent manually-selected topics and sentiments that match a specific field.
  • a_topics_sentimnts_original_for_[field] — tbd
  • a_topics_sentiments_tagged_after_manual_updates — A comma-separated list of the topics that match a record, combined with the sentiment matching the same regions. This field lists topics originally discovered by Experience Cloud and those tagged manually by users.
  • a_topics_sentiments_tagged_original — A comma-separated list of the topics that match a record, combined with the sentiment matching the same regions, as originally discovered by Experience Cloud (excluding topics tagged manually by users).
  • a_topics_tagged_after_manual_updates — A comma-separated list of the topics that match a record. This field lists topics originally discovered by Experience Cloud and those tagged manually by users.
  • a_topics_tagged_original — A comma-separated list of the topics that match a record, as originally discovered by Experience Cloud (excluding topics tagged manually by users).
  • a_topics_with_hierarchy_for_[field] — A comma-separated list of the topics and subtopics that match a specific field.
  • a_topics_with_manual_sentiments_for_[field] — A comma-separated list of the manually-selected topics and sentiments that match a specific field.