I’m very excited to announce the newest addition to the HumanListening platform – Text Auto Reports. This new feature uses AI to generate insights from Conversational AI chats, or open-ended data, in just a few short clicks.
We have been working with AI for years, first debuting in EVE, our Conversational AI chatbot that interviews respondents to get deeper insights than traditional surveys. In the same way Conversational AI has changed what is possible with gathering qualitative data, Text Auto Reports sees a significant leap forward in analysing qualitative data. This has been built to streamline the process of getting insights out of text data and provide deeper insights than previously possible.
What are Text Auto Reports?
Text Auto Reports combine a series of AI-based functions to produce a comprehensive end-to-end analysis of qualitative data. This allows users to quickly go from raw text data to actionable insights, with minimal effort. It achieves this by merging Natural Language Processing (NLP) tasks (including sentiment analysis, topic extraction, text classification & summarisation) with established analytics processes to provide analysis that you can rely on.
The reports are a combination of visualisations and natural language. You can expect to see striking charting of the insights, as well as sharp, to the point summations. Your data is automatically tagged into topics & subtopics, with charts showing you which subtopics are the most prevalent, as well as which are more positive / negative in your respondents' eyes.
To support even greater depth, you can click on a subtopic in any of the charts to quickly see a summary of the texts that have been tagged, as well as example comments. We recently added a new feature that highlights exactly which parts of a conversation relate to a selected subtopic, making it easier to identify key subtopics and trends within your open-ended data.
The below explores the subtopic of ""customer service"":
Text Auto Reports were built to handle both Conversational AI chats and standard open-ended data, so no matter the format of your text data, you can get to real insights quickly and easily!
Saves you time while improving quality
Text Auto Reports utilises recent AI advancements to streamline the process of analysing text data, while improving accuracy when compared to current classification techniques. You can now go from raw text data to actionable insights in a matter of minutes. The report will also dynamically update as you apply filters etc., so that you won’t waste time running different versions of the reporting.
Alongside the release of Text Auto Reports, we are also introducing Auto Classification in the HumanListening platform. This feature takes your text data, uses AI to develop a topic model, and then classifies your text against this model (including tagging sentiment at the sub-topic level). It is demonstrably accurate, classifies your data into meaningful topics without any manual effort and does so with almost no setup time. Auto Classification is available in both Text Auto Reports and in our interactive analyst workspace, Text Explorer.
Applications for Text Auto Reports and Autoclassification
- Accurately code 1,000s of comments in a few seconds
- Run multi language analysis across multiple markets
- Identify key drivers of satisfaction in a CX study
- Identify drivers of consumer tension on your NPD study
- Explore drivers of brand preference on your brand tracker
- Test three branding concepts and understand preferences to create an even better fourth one!
How to get started
Getting started is quick and easy:
- Navigate to the Text Explorer app in HumanListening.
- Hit the + (Create New) button, and select ‘Text Auto Report’.
- Select a data source and choose the text variable(s) to analyse (e.g. an EVE conversation or open ended question). A separate report page will be created for each selected text variable.
- To ensure Auto Classification and summarisation are relevant to your project goals, enter a brief description of the report’s context.
- Apply any relevant filters to drill down into results.
- View and share your report via the tile in the Projects section in Text Explorer and/or export the report directly into PowerPoint
Future of AI in Customer Insights
There are big plans for Text Auto Reports, with several features and enhancements already in the pipeline. This includes:
- Auto Classification Model Review: Introducing a step for reviewing the model generated by Auto Classification before tagging (so that topics can be added, removed, or edited)
- Comparisons Section: Allowing users to see which subtopics are more prevalent between two groups
- Trended Data: Further support for showing insights over time
- Topical Filtering / Deep Dive: If a topic is of particular interest, a user can run a report focusing on just that topic (either before a report is run or after a report is already generated).
These are just a few of the features you can expect to see in the coming weeks and months. At HumanListening we’re dedicated to ensuring you get the most out of your data, and we are continually looking for ways to use AI to shape the future of insights.
Text Auto Reports & Auto Classification allows insights to be quickly gleaned with minimal effort, while also offering the flexibility to work across a range of different sources & data types. Combined with other capabilities available within platform, such as EVE, it has never been easier to go from business question to data-driven answers.
Want to learn more?
If you’re an existing customer looking to get additional information on using Text Auto Reports, check out our help guide or reach out to your account manager directly.
If you’re interested in learning more about Text Auto Reports, or the HumanListening platform more generally, simply get in touch with us below.
Get in Touch
