Example Questions for Answers

Answers has a carefully built system to make sure your questions are interpreted correctly and produce the most useful responses. As we improve this system over time, more and more questions will be better supported, so Thematic can give you a quality answer. Here are some example questions that Thematic currently supports. Getting started with asking questions on Answers can be simplified with a few clear examples and use cases. These examples will provide insights into the types of questions we support and how to troubleshoot effectively if you encounter any issues.

To learn more about Answers, visit this help article.

1. Explain one aspect of your data

You can ask questions to receive a summary of your feedback based on various parameters such as date range, category, theme, sentiment, or cohort. You can apply these using the filters, and Thematic may be able to interpret filters from your question Some examples of questions for this purpose include:

Ask about a theme or specific idea

What do customers say about the price?
What do customers say about the process of getting a refund?

Ask about a particular segment by naming the filter in your question or applying a filter

What do customers mention the most in their feedback?
What do Detractors mention the most in their feedback?
What do Detractors in October mention most in their feedback?

Use the positive/negative sentiment and category filters to get more specific

What do customers dislike about refunds?
What do customers like about refunds?
What questions do customers have about refunds?
What issues do customers have with refunds?
What requests do customers have about refunds?
What are the most common issues people face?
What do people like the most?
What do our highest paying customers request?
What were the issues with customer service last Christmas?

Ask about counts of users or survey responses

How many customers mention price?
What percentage of comments mention refunds?
Quantify how many issues were related to onboarding.

2. Compare two segments and get a summary

You can define and ask Thematic to compare two segments. Define the two segments in your question, any filter you apply will be applied to both. Examples of questions for this use case include:

Compare promoters to detractors.
Compare this month to last month.
Compare positive to negative feedback.

3. Get a score about your feedback quickly

You can quickly ask about a metric or score for a specific dimension without the need to sift through extensive data. Some example questions for this use case

What is the net promoter score for June?
What is the NPS for the California customers?
Which city has the highest NPS?
Which city has the highest CSAT?
Can you show the performance on the Boston Branch?
Which RSP had the highest NPS?
What city do people like the best?
Which month had the most positive feedback?
What is the NPS score by state?

4. See what’s caused a change in your score

You can investigate the factors contributing to a change in your score over a particular period. Some example questions for this purpose are:

What caused the NPS to change between June and May?
What has caused the score to decrease compared to last month?

Unsupported Questions

If a question you asked is not yet supported, consider either rephrasing your question to align with our current supported question types or help us improve the way we categorize questions by providing feedback. Any questions that you ask are tracked by Thematic to help us improve answers, so don’t be afraid to try a question to see if Thematic is able to handle it.

Not Enough or No Data

If you selected a single data set and the question you asked is supported but has less than 10 rows of feedback, you will receive a "not enough" or "no data" result. In this scenario, consider widening your search by adjusting your filters or providing us feedback if you suspect missing data.

If you selected multiple datasets for your question and at least one contains relevant data, you will see the answers for each individual dataset where applicable.