Idiomatic is committed to helping teams better understand customer feedback. Of course this means we’ve also been collecting feedback ourselves! With this major update, clients will see that we’re listening to all of that feedback. There are several exciting changes that are about to go live.
Tl;dr: real-time analysis is here!
One of the features most frequently demanded by our clients was to deliver insights in real-time. With this update, we will begin analyzing data as soon as it comes into a client’s helpdesk. Our Machine Learning models will predict categories, issues, sentiment, and confidence scores for all new tickets just a few minutes after they are created in the helpdesk. The tickets will appear almost immediately in the Idiomatic dashboard.
Better organized filter menu
We’ve taken the opportunity to revamp one of the most highly used aspects of our product: the filter menu.
You will see the new menu is organized into 3 sections:
- Issue Tags
- Sentiment Tags
- Custom Filters
Issue tags will contain filters for the issue tags which we discuss below, sentiment tags will contain filters for sentiment as discussed below, and custom filters will contain the rest of the filters which are derived from client side filters or fields in the client data source.
More robust sentiment models
We’ve re-organized the way in which we analyze sentiment to make it more robust across various data sources, handle increased volume of analyzing all client tickets as they come in, and be simpler and easier to understand.
The new sentiment categorizations are:
Comment type revamped as issue type
Many clients have requested greater visibility and control over classifications. With this release, users with admin privileges will be able to annotate each issue as one of the following:
- Request for Action
- Technical Problem
Existing issues will come pre-annotated based on how Idiomatic has been classifying them. But admins will be able to change these classifications of each issue at any time by clicking on the new tab Categories and Issues.
We’ve eliminated Complaint and Compliment as their redundancy with Negative and Positive Sentiment caused confusion.
Flagging tickets for review
We’re introducing a ticket flagging tool to the dashboard. For any ticket you can click on the upper right to flag it.
You will now be able to mark tickets for review by the account analyst and also leave a note as to what is wrong or needs to be reviewed.
When can you get it!
These updates will be rolling out live to clients over the next few weeks. We are extremely excited to see how these changes allow our clients to get insights faster with more precision. We would love to hear any and all feedback at firstname.lastname@example.org.