A tag is a label or identifier that helps to categorize tickets into different groups. They facilitate better ticket management and clearing the triage by allowing the support team to focus on tickets that require immediate attention.
For example, tickets that are tagged as 'cancellation', 'modification', or 'payment failure' can be marked as high priority to facilitate quick resolution. While tickets can be tagged manually by agents, it can become tedious and challenging when the volume is high. Reading each ticket and identifying the relevant tags can also pose higher chances of error, due to a lack of standardized approach, leading to different agents using different tags for the same category. This can also potentially impact the SLAs. This is where AI-based Auto Tags can be useful.
Availability
Permission Required
Users with administrative privileges can enable Zia and Zia Auto Tag.
Zia can train itself by analyzing existing tickets. By reading the descriptions of each ticket, it identifies a group of similar keywords and forms a cluster. It then generates a suitable tag name for every keyword cluster that is created. When new tickets are created or submitted to Desk, it analyzes them, identifies the keywords, matches them to a relevant keyword cluster, and associates the tag with the tickets.
As the ticket conversation progresses, Zia identifies new keywords and continues auto-tagging them according to their relevance, without deleting or replacing the previous tags. This repeats for each customer reply. Hence, a ticket may have multiple tags associated to it, which helps the agents classify them into appropriate groups.
For example, if a ticket is created for a faulty product, it will be tagged as 'product issue'. Later, if the customer requests a return and refund, Zia will add another tag as 'return and refund' while retaining "product issue". This will give the agent context for the ticket, that the return and refund was requested because the product was faulty.
These tags can also play a major role in identifying the number of issues that are reported for a particular product. Businesses can further investigate the issues to minimize the complaints and improve customer experience.
- A department must have a minimum of 3000 individual tickets via email, help center, web forum, community, or social for Zia to train and generate auto tags.
- Tickets created via Chat (SalesIQ, IM), Phone, and Custom Channels will not be considered by Zia for training.
- Zia auto tag is department-specific and can be enabled based on the requirement.
- Auto tags will be generated for tickets created through the following channels: emails, help center, social, web forms, socials, manually created tickets, and tickets created via API depending on the channel.
- Zia will only auto-tag tickets that contain keywords that match with the keyword cluster.
Enabling Zia Auto tag
Users with administrative privilege can enable Zia and Zia Auto Tag. Note that before enabling auto-tag, you must enable Zia for the Desk account. Enabling Zia is an organization-wide action, which will apply across departments and portals. However, you must enable Zia Auto Tag for every department individually for it to work.
To enable Zia auto tag
Log in to the Zoho Desk account with administrative privileges.
- Go to Setup > Zia.
- Toggle Zia to enable all Zia features.
- In the Predictions tab, toggle Ticket Auto Tags.
The tags will be associated to the tickets and can be viewed by the users under key information in the ticket properties. Within the ticket detail page, under History, you can view when the ticket was tagged.
Tag under key information
Note that under History, the system will always display that the admin has tagged the ticket, even though the action is carried out by Zia.
The auto tags provided by Zia can be modified or deleted by the users if needed. For example, instead of "product issue", if you want the tag to be "defect", you can edit the tag.
Note
- Auto tags can be edited or deleted to your preference. The keyword cluster is generated after analyzing the existing tickets from the database, and cannot be edited.
- The Manage Auto Tag option will only be displayed when the department has more than 3000 tickets. Zia Auto Tag can be enabled even if there are fewer than 3000 tickets, but the auto-tagging and keyword generation will take place only when there are more.
- Go to Setup > Zia.
- Go to Ticket Auto Tags and click Manage Auto Tags.
- Click on a tag from the available list.
- Click Change Tag Prediction to edit the tag.
- Enter a tag name or choose a tag from the available list.
- The selected tag will be considered the new tag.
- Click Delete Tag and select a tag.
- Click Delete in the dialogue box to confirm.
Training Zia to improve auto tagging
Zia improves its analytical abilities by learning continuously from the available data. In some cases, when the suggested tag doesn't match the business's requirement, that will need to be edited or deleted, so that Zia can utilize these modifications for its training purpose.
This continuous training will help Zia understand the preference and assign tags to one's unique requirement.
Note: If agents manually add tags to the tickets, those will not be considered by Zia for training purpose. Only the auto tags that are edited and deleted by the user will be used by Zia for training. To improve auto-tagging
- Go to Setup > Zia.
- Under Zia Predictions, click Manage Auto Tags.
- Click Set Auto Tags from the bottom panel.
- Select a relevant tag from the given options or add a tag for the given keyword cluster.
If a specific tag is used more frequently, it indicates that the business is receiving more queries for a particular requirement or issue.
For example, if 20 tickets are tagged as "fraud scam", the business should actively follow up with the team to understand and address the underlying issue.
The frequently occurring tag will be considered a trending tag by Zia. The top trending tag will be displayed in a bigger font for easy identification. The underlying sentiment of the tag will be displayed based on Zia's analysis of the ticket conversation. With the help of these insights, businesses can identify the problem areas and take quick action.
Zia will display the trending auto tags in the dashboard. It will display:
- The tags that have been trending for the last 24 hours.
- The dashboard can be viewed either in a cloud format or tabular format by setting the filter.
- Hovering on each tag will show the number of tickets that are associated with it.
- The trending tag can be identified with the font size. A smaller font indicates a less common occurrence.
- The sentiment of each tag will be displayed.
- Go to the Analytics tab.
- Under Dashboard, click Zia Dashboards.
- Click Predictions Dashboard.
Auto tags can also be used to trigger workflow rules. For example, if the organization wants to alert the Security team about fraudulence or scams, then the tickets that are tagged as "scam", "fraud", "transaction fraud", and so on can be assigned to the respective team automatically for immediate action.
These tags can be set as criteria in workflow rules.