All Collections
CXInsight Advanced Analytics
Understanding Tags in CXInsight
Text Analytics | Smart tags by Category
Text Analytics | Smart tags by Category

Smart tags are applied to feedback using machine learning for out-of-the-box topic auto-categorization in CXInsight.

R
Written by Ryan Smith
Updated over a week ago

Smart tags are themes that are automatically applied to feedback for various industries and types of feedback. These tags are based on our pre-trained, proprietary algorithms. Smart tags correspond to a set of topics common to an industry or type of feedback. Note: algorithms also designate sentiment of each comment and each topic within the comment, as in the employee feedback example below:
​

Smart Tag Categories

  • E-Commerce

  • Employee Feedback

  • Online Food Delivery

  • SaaS Dev Ops

  • SaaS Social Media

  • Support Case Feedback

  • Politician Feedback

E-Commerce

E-Commerce analyzes feedback from users of any sites that have e-commerce components. Online shopping platforms share many similarities that are reflected in what we believe to be general tags for this use case. The supported tags and their descriptions are:

  • Billing - Any feedback having to do with the payment process. Comments about declined credit cards or double charges would fall under here.

  • Customer Service - Applied to comments detailing experiences with customer service for the platform.

  • Customer Suggestions - Any suggestions that a respondent may have about enhancements to the platform, or general notes on areas that can be improved.

  • Discounts - Applied to any comments mentioning either the benefit or lack of discounts and deals for the platform.

  • Fitting - Any feedback having to do with the "fit" of the product. Note, this tag mainly applies to clothing retail.

  • Inventory - Any feedback that pertains to the selection or amount in stock of a product(s). Example feedback would be "Love the variety!" or "X was sold out".

  • Order/Checkout - Applied to comments mentioning the ordering and/or checkout process. Examples of this include "It was easy to add items to my cart", or "The checkout window was very clunky".

  • Price - Any comments that mention the relative cost of products, i.e. too expensive or relatively cheap. Also contains comments mentioning the value of a product for the money spent.

  • Product Quality - Applied to feedback that mentions the quality of the product. This can be hard to capture for general purpose products (ask us how to train smart tags unique to the products you sell!), so this tag will typically pick up general terms applied to products like "sturdy", "breaks easily", or "works great".

  • Returns - Comments pertaining to the return process. Most often these comments will make explicit use of the words "return" or "refund".

  • Shipping/Packaging - Comments that mention any aspect of the shipping process, or the packaging of the product. This can range from "I've waited 4 months for my order" to "The lamp was broken when it arrived".

  • Subscription - Feedback that has to do with a subscription, especially either renewal issues or cancellation. For single time payment platforms, this tag will not apply.

  • Technical - Any comments that pertain to the UI or technical experience with the platform. These can range from commenting on good search functionality to mentioning common bugs they experience while shopping.

  • Value Prop - Applied to feedback that mentions value propositions of the product. This tag is applied frequently and mainly corresponds to comments that highlight or compliment a specific aspect of the product or platform.

Employee Feedback

Employee Feedback categorizes feedback from employee surveys and reviews. The supported tags and their descriptions are:

  • Basic Compensation - Feedback that is about pay rates (hourly, salary or commission)

  • Benefits - Applied to comments that mention work-related benefits (i.e. medical insurance, 401K, PTO, etc.)

  • Bonus Perks - Feedback that is about additional job perks, such as free meals, bonuses, and other incentives.

  • Career Growth - Comments that concern career development. These can include phrases like "promotion" or "career path".

  • Challenging Work - Comments that focus on the difficulty or engagement level of work.

  • Culture - Related to office culture - i.e. "nice people at the office", or "work environment is not great".

  • Day of the Week - Feedback that mentions a specific day of the week.

  • Direct Manager - Feedback pertaining to a direct manager or boss.

  • Diversity - Feedback related to diversity and inclusion.

  • Engagement Valued - Tagged on comments that mention job engagement or feeling valued by the company. Examples may include "I love my job!" or "I have been feeling ignored by management lately".

  • Facilities - Applied to feedback that pertains to work buildings or environment. Common examples include "Cubicle feels tight" or "Office is in a great location".

  • Health & Wellness - Any feedback that mentions physical or psychological ailments the employee might be experiencing.

  • HR - Feedback that mention HR. This tag is fairly simple, and usually just applies to feedback containing the phrases "HR" or "Human Resources".

  • Leadership - Comments that pertain to leadership shown by bosses or executives.

  • Management - Any feedback that mentions management or managerial styles. Some examples could be "I don't enjoy the constant micromanaging", or "I'm glad management is so hands on".

  • People Skills - Feedback that specifically mentions "people skills".

  • Training - Comments related to employee training.

  • Weather - Comments that mention the weather.

  • Work/Life Balance - Feedback that has to do with an employee's stress level and/or work life balance. Examples could include "My home life is suffering from spending so much time at the office", or "My vacation was wonderful, and I'm ready to get back into work!"

  • Work Remote - Comments that specifically pertain to working from home or other remote locations (not in the office).

  • Workload - Applied to feedback surrounding the topic of employee workload. Common elements of feedback tagged by Workload include "stressed with work lately", "workload feels really light" or "been bringing work home with me too often".

Online Food Delivery

Online Food Delivery is fairly straightforward. This category deals with any tags you would find associated with a food delivery service, specifically with an online platform. The models were trained on feedback from customers receiving food, commenting on their experience with the service. The supported tags and their descriptions are:

  • Customer Service - Applied to comments detailing experiences with customer service for the platform. This can include calls, emails, or chats through the platform.

  • Delivery - Any feedback pertaining to the delivery process itself. Most often, these comments mention the wait time until receiving the food.

  • Food Choice - Applied to comments on the variety of food available on the platform. Usually has to do with the list of restaurants, or menu of a specific restaurant.

  • Food Quality - Comments that mention the quality of the food delivered. Comments relating to the food being cold will most likely also be tagged with Delivery, due to the implied long wait time.

  • Order/Checkout - Feedback that mentions the food ordering or checkout process. This can include comments like "It was easy to reorder the usual" or "I couldn't select the correct choice of meat".

  • Payment - Any feedback having to do with the payment process. Comments about declined credit cards or double charges would fall under here.

  • Price - Comments that concern the price of the service or order.

  • Promos/Coupons - Applied to any comments mentioning either the benefit or lack of discounts and deals for the platform (including promotions).

  • Reviews - Any feedback related to reviews of menu items or restaurants.

  • Technical - Comments on the technical aspect of the platform, such as the UI or performance. This could include feedback like "I wish you could search for a specific dish", or "The app is always really slow".

  • Value Prop - Applied to feedback that mentions value propositions of the product. This tag is applied frequently and mainly corresponds to comments that highlight or compliment a specific aspect of the product or platform.

SaaS Dev Ops

SaaS Dev Ops models were trained on a dataset of SaaS companies used primarily for improving the development experience. However, this is also our largest dataset, and we have found many of the tags apply even to feedback not pertaining specifically to Dev Ops platforms. Because of this, we have made it our go-to category for SaaS company feedback. The supported tags and their descriptions are:

  • Alerts/Notifications - Any feedback that mentions alerts, notifications, popups, or spam.

  • Bugs - Feedback that comments about encountering bugs on the platform. Some examples might be "The service keeps disconnecting", or "My dashboard is frozen".

  • Customer Support - Applied to comments detailing experiences with customer service for the platform.

  • Documentation - Feedback that mentions any relevant documentation on the platform, including tutorials, walkthroughs API docs, etc.

  • Feature Request - Any suggestions that a respondent may have about enhancements to the platform, or general notes on areas that can be improved.

  • Onboarding - Feedback related to getting started with the service. This commonly references experiences with setup or installation of the service.

  • Performance - Feedback that pertains to technical performance of the service. Comments can include "The app loaded really fast!", "The UI was slow to respond", or "Too much downtime with this service".

  • Price - Any comments that mention the relative cost of service packages, i.e. too expensive or relatively cheap. Also contains comments mentioning the value of a product for the money spent.

  • Reporting - Applied to feedback concerning any report functionality of the service. This can include analytics, charts, graphs or other analyses.

  • UX/UI - Feedback that mention the user experience, or the UI in general. Typical comments include "App was easy to use!" and "I wish the navigation was a little more intuitive".

  • Value Prop - Applied to feedback that mention value propositions of the product. This tag is applied frequently, and mainly corresponds to comments that highlight or compliment a specific aspect of the product or platform.

SaaS Social Media

SaaS Social Media applies to feedback from users of SaaS platforms that deal specifically with social media. Our training set includes comments mainly from social media management platforms, and relate to users' experiences there. The supported tags and their descriptions are:

  • Analytics - Feedback that pertains to any type of reporting or analysis shown to the user.

  • Feature Request - Any suggestions that a respondent may have about enhancements to the platform, or general notes on areas that can be improved.

  • Performance - Feedback that pertains to technical performance of the service. Comments can include "The app loaded really fast!", "The UI was slow to respond", or "Too much downtime with this service".

  • Price - Any comments that mention the relative cost of service packages, i.e. too expensive or relatively cheap. Also contains comments mentioning the value of a product for the money spent.

  • Scheduling - Applied to feedback that mentions any aspect of scheduling. This is more specific to the social media aspect than other tags, and can pick up feedback like "I want to be able to plan posts on multiple platforms" or "It's great being able to coordinate a Facebook post so far in the future".

  • UX/UI - Feedback that mentions the user experience, or the UI in general. Typical comments include "App was easy to use!" and "I wish the navigation was a little more intuitive".

Support Case Feedback

Support Case Feedback deals with responses to surveys conducted after (or during) interactions with customer support. While our data primarily trained on customer support cases that had been closed, much of the language applies to any surveys conducted for ongoing support cases as well. The supported tags and their descriptions are:

  • Competency - Any feedback pertaining to the skill level of the support agent. This includes attributes like "knowledgeable", "helpful", or "experienced".

  • Escalation - Feedback that mentions a case being escalated past the original support agent. Typically this involves dealing with a senior support rep or manager to solve the issue.

  • People Skill - Any comments that mention how pleasant the interaction experience with the agent was. Examples include "<Agent Name> was so friendly!" or "<Agent Name> did not seem very professional".

  • Promptness - Comments that mention the response time by the service agent/rep in support interactions.

  • Repeat Info - Applied to feedback that mentions being told the same info repeatedly. For example: "I told the agent that I had already installed the product. However, they then repeated the installation instructions for a third time".

  • Resolution Time - Applied to feedback that talks about the length of time for a case to close, and their support issue to be resolved. This can sometimes overlap with Promptness, as a long wait for a response will typically also yield a long resolution time.

  • Suggestions - Any suggestions that a respondent may have about enhancements to the platform, or general notes on areas that can be improved.

  • Unresolved - Applied to feedback that says the original support issue was never resolved. For example: "I called because I could not log in, but the issue was never fixed".

  • Updates/Followup - Any feedback that mentions updates from the agent, or the back and forth followup as the issue is being addressed. This could include comments like "<Agent Name> was very communicative throughout the process", or "I never received a call back after being told they would look into the issue".

Politician Feedback

Politician Feedback deals with responses to surveys that gather feedback and sentiment regarding candidates or parties. The supported tags include Leadership,

Competency, Character, Behaviour, Policy. Additionally, rules-based tags categorize feedback related to Economy, Immigration, Healthcare, Foreign Affairs, Safety & Security, and Race Relations.

Did this answer your question?