EN

Moving the Sentiment Analysis Further: Introducing an aspect-based analysis

We never cease to improve the functionality of our system, and it’s time to announce a new feature! YouScan team has developed an aspect-based sentiment analysis.

Valeriia Zhyvotova

Content Marketing Manager

02 April 2020

presentation

In a nutshell, aspects are characteristics of the product/service to which the users can express their attitudes. Aspect-based sentiment analysis is a finely-crafted model that allows brands to automatically break down customer feedback, define the public’s opinion, and allocate the aspect on a sentiment level. You can confidently rely and trust the data because the accuracy of defining the aspect is 85-90% while the accuracy of defining the sentiment of the aspect is higher than 95%.

While the fundamental sentiment analysis provides you with the overall sentiment of the monitored topic, a new feature allows you to see which particular aspect of your product/service generated this attitude. Therefore, you can see individual sentiment towards your product taste, price, service, quality, and other popular characteristics customers discuss.

Moreover, YouScan provides you with an opportunity to apply various filters and dive deeper into details to see how aspects vary depending on the geo, gender, age, language, and other filters. Aspect analysis allows you to instantly identify potential risks for your business and take action straight away. However, you can also discover your strenghts and what customers love about your brand.

Nailing down the benefits of the sentiment analysis: what does a brand win?

presentation

1. A brand gets closer to customer opinion, finds insights, and boosts the customer-centric experience

Usually, when analyzing sentiments of product or service reviews, you’ll want to know which particular aspects or features people are mentioning in a positive, neutral, or negative way. You want to see customers’ pain points and instantly address them. A new feature allows you to choose one particular aspect, hone in on the details, and get a complete picture of your customers’ opinions, expectations, and complaints.

A convenient interface of sentiment analysis provides brands with an opportunity to detect trends and find insights around one particular aspect. Also, you can apply various filters and play with visual insights, auto-categories (WOM, articles, promotions, etc.) to uncover ideas and inspiration for your next marketing campaigns. Sometimes you have to look beyond your brand and get to know your audience intimately

2. Analyzes the competitor’s strengths and weaknesses

The aspect analysis helps you to compare your brand’s features with competitors’ ones. You can detect the flaws in your product and work on improvements. Moreover, you understand your strengths and get insights on how to promote them more efficiently.

3. Prevents or deals with a potential crisis effectively

A spike in negative sentiment around a particular aspect requires a prompt reaction. Now, you can apply various filters to find the root of the problem and create an emergency plan to fix this negativity. Without aspect analysis, it’s more challenging to detect the reason for the spike, which leads to a delay in the solution of the issue and can undermine customer loyalty and brand health.

4. Saves time and organizes data

The raw collected data from social networks is unstructured and unorganized. You get vast amounts of mentions from chats, social media conversations, articles, etc., but it’s hard to analyze and sort through, not to mention that it’s time-consuming and expensive.

One person can analyze and categorize by aspects around 500 mentions per day. Meanwhile, the system can analyze millions of mentions in a day automatically tagging them with aspects.

How does it work?

Let’s dive into the particular case to understand the full potential of the new feature.

Imagine you’re Starbucks. In the analytics dashboard, you click on the aspect tab and find the sentiment by aspect graph. Here, the first 5 aspects on the radar chart are related to the highest percentage of mentions where Starbucks was discussed by its target audience. Let’s get into the finer details.

In the time period from 12.02.2020 to 12.03.2020, taste, price, service, design, and quality were the top 5 aspects of Starbucks that people talked about.

presentation

It’s no wonder that Starbucks has a high percentage of positive mentions, especially in regards to taste, design, and quality but every business is interested in improving its product. Thus, it needs to know not only the praising feedback but also customers’ worries, complaints, and suggestions.

The chart reveals that people complained most about service (26% of negative mentions) and price (20%). The new feature allows you to discover what exactly caused this negative attitude.

Simply filter out the mentions by negative sentiment and choose only those aspects with high negative sentiment.

presentation

Afterward, you can see what sparked negative sentiment among Starbucks customers by filtering out these particular aspects. Some customers were unhappy about the overpriced coffee and the taste.

presentation

Final thoughts

YouScan is striving to be better for our clients and we’re working hard on honing our features to empower brands with unique social listening capabilities. YouScan is already offering more services than a simple social media monitoring tool - with a new feature you can conduct research that generates qualitative and quantitative results. We believe that aspect-based sentiment analysis is another victory of our team and the success of your brand analysis!

Still don’t monitor your brand on social media? Request a YouScan demo now!

Subscribe to our Blog