Syncly is an AI-powered customer feedback analysis platform that helps identify key customer issues by analyzing various text data such as chats, surveys, and reviews. We have recently launched Insight beta version, enabling a more holistic AI analytics through 'Semantic unit analysis', making it easier but more thoroughly grasp customer insights.

What is Semantic Unit Analysis?

Semantic unit analysis is an analytical technique that extracts content related to the theme from the text. It is a more segmented and advanced content analysis method than the existing method of analyzing text at the sentence or the message level. All in all, it captures the details otherwise could have overlooked by conventional analytics method.

As shown in the image above, the original text-based review data can be divided into several semantic units. This is because multiple themes almost always exist within a single review or a conversation. Therefore, if you analyze the semantic unit by themes, it becomes easier to quickly grasp the key content of each theme, discovering the detailed insights from the review.

How we structure Insight

Through the Insight in Syncly, users can utilize more sophisticated AI content analysis through Semantic unit analysis. When you select a theme, mentions (content relevant to the theme from customer review) are automatically extracted to roll up to issues, then to themes in a three-depth structure by clustering based on commonality.

For example, if you add the theme of "fabric" to analyze fashion-related customer reviews, it extracts all contents related to the theme. Characteristics of the fabric(e.g. texture, feel, or quality of the fabric), the thickness of the fabric, and content related to laundry (e.g. how well the fabric holds up after washing, whether it shrinks or fades in color) are extracted and clustered. By linking these traits, issues are formed as "Fabric Thickness Concerns”, “Fabric Composition Inquiry”, “Fabric Durability Concerns”, and by showing the semantic unit text together, the content can be easily understood at a glance.

How to leverage Syncly Insight

Simply by adding a theme on Insight page, Syncly will automatically extract the main content from the existing data in semantic units and provide the results by clustering issues.

(1) Adding a Theme

There are three ways to add themes. You can directly them, or get additional themes recommended by AI, recommended theme by industry. Based on the selected themes, AI runs semantic unit analysis to extract Mentions(semantic unit) and Issues(clustered semantic unit).

• AI Suggestions: This is the Themes recommended by Syncly.

• Templates by Industry: This is a Theme template recommended based on how our customers build taxonomy by industry.

• Type the Theme: Users can create a Theme by directly typing in the input box.

(2) Results of AI Analysis - Theme list, Issue List

Check the results extracted by the Theme to easily identify customer issues and pains. Not only the semantic unit analysis, but also different levels of positive and negative sentiment can help users identify urgency of each issue.

For example, you can assess the significance by monitoring the number of issues and mentions in the ‘Material' related theme. Within this theme, we extracted 7029 mentions and identified 24 distinct issues. A deeper exploration of the 'Material Thickness Concerns' issue reveals 10 positive mentions and 100 negative mentions.

(3) Ask Syncly AI for suggestions

You can utilize Syncly AI by tapping ‘Ask for an action’, ‘Ask to summarize’, and more. Users can enter any questions to discover additional insights.

• Ask for an action: Syncly AI suggests what action items can be helpful in solving the issues for various functions—customer experience, product, and for operations.

• Ask to summarize: Syncly AI summarizes and provides every detail of each issue.

Insight Demo Video

Curious about Syncly's latest Insight feature?

Book a demo to receive 1,000 free data analysis credits and a free trial to analyze your customer feedback data!