What, how, why? – How our Quantified Sentiment Insights will transform decision-making

Unlock the power of customer insights with Gavagai Explorer, our state-of-the-art tool that not only identifies what your customers are talking about but also reveals how they feel and why they hold those feelings.

How many hours are you still spending on insights generation? We know that even with the help of tools such as Gavagai Explorer, this can be a time-consuming and tedious task. Now imagine having actionable insights generated for you automatically.

Our newest feature, Quantified Sentiment Insights, is powered by state-of-the-art large language models (LLMs) and allows you to extract both positive and negative sentiment reasons and insights from your data, quantified by how many respondents are affected by them. 

What are Quantified Sentiment Insights?

Quantified Sentiment Insights, or QSI for short, is the newest feature here at Gavagai, and part of Gavagai Explorer, a B2B SaaS text analysis tool for the Customer Experience domain. With Explorer, you can analyze unstructured text data such as surveys (NPS, CSAT), support tickets (email, chat), service or product reviews, and social media.

Let’s have a look at what this means for you.

Customizable modeling

Topics describe what your customers are talking about. Whether you prefer to mold the topics for your insights yourself or seek assistance from Gavagai Explorer, our platform empowers you to tailor the analysis to suit your unique business needs. With intuitive tools such as automatically generated term suggestions or rule-based logic features, you have the flexibility to refine your models and extract insights that resonate most with your objectives.

Unmatched insight generation

After deriving the most important topics from your texts, it is time to generate insights with the help of a state-of-the-art Large Language Model. Insights are organized by sentiment, summarizing how your customers are feeling about a particular topic. They are presented to you in a user-friendly dashboard, making it easier than ever to uncover hidden patterns and trends within your data.

Actionable intelligence

Beyond identifying topics and sentiments, Quantified Sentiment Insights provides you with actionable insights backed by solid evidence to show you why your customers are holding these feelings. From pinpointing areas for improvement to highlighting untapped opportunities, each insight is designed to empower you to make informed decisions that drive tangible results.

QSI can provide insights such as: 

122 of the guests at the hotel are unhappy with the room because they find it to be smaller than advertised.
On the other hand, 337 of the guests liked the room since it was well maintained and clean.

5% of the tenants are unhappy with the temperature in their apartment since they think it’s too cold.
This is an improvement by 3 percentage points since the last quarter.

1352 of the customers express appreciation that their orders are delivered quickly. This creates a positive customer experience and increases the chance of repeat purchases.
In comparison, only 139 customers express dissatisfaction when their orders are not delivered within the expected time frame.

Transparency is key

By keeping track of the number of respondents that have contributed to an insight, Gavagai Explorer provides you with a ranked list of insights, reflecting the importance of them in the view of your customers (the more support for an insight, the more important it is). 

Combining the analyst’s domain knowledge and the automatically produced Quantified Sentiment Insights, the analyst can assess the feasibility of addressing each insight as part of an action plan.

Since each insight comes with the texts that express that particular insight, it is easy to verify for you that the insights are valid.

How does it work?

For a more technical perspective, we have asked our resident expert Fredrik Olsson, Head of Data Science at Gavagai who holds a PhD in Computational linguistics and 25+ years of industry experience, to provide you with a short explanation about the inner workings of what it takes to produce Quantified Sentiment Insights:

“The process leverages Gavagai’s proprietary technology to identify topics in text, with state-of- the art Transformer-based classifiers fine-tuned on select training data to classify utterances with respect to their sentiments, and a multi-stage LLM prompting approach for generating the insights. The Gavagai Explorer runs on our private cloud located in Stockholm, and the LLM used is GPT-4 Turbo deployed in Microsoft Azure, also in Sweden.

I’m very excited about the launch of QSI! The feature has already received a lot of praise from the customers that have had early access to it.”

To find out more about how to access QSI inside Gavagai Explorer, be sure to have a look at our documentation where you will find all the necessary information to try it out for yourself.

Why does it matter?

Because time and quality matters:

  • A transparent and trustworthy AI solution will save you hours of precious time and maximize the results of your customer experience analysis.
  • Quantified Sentiment Insights go beyond surface-level analysis. By understanding not just what your customers are talking about, but also how they feel and why those sentiments matter, you can make informed decisions that drive tangible results.

Get ready to unlock a whole new level of understanding from your text data and transform the way you make decisions.

Want to try this kind of analysis on your own data?

Gavagai Explorer reveals why your customers feel as they do.

Get in touch!

This website uses cookies to ensure you get the best experience.