Customer surveys, reviews and support tickets hold insights that can drive retention, increase conversion rates, and improve customer lifetime value. In this guide, we'll teach you the ins and outs of customer sentiment analysis so you can turn feedback into actionable customer insight.
Customer feedback can come from many different channels. Whether it's NPS or CSAT survey scores, social media comments, or data logs from email, live chat and phone customer service.
The most powerful customer feedback is qualitative. For example, free text fields in customer feedback surveys or chat logs from customer support conversations. However, at scale, qualitative feedback is hard to measure. You'll need a robust data tagging methodology, lots of time, and an objective eye.
A handful of interviews are second best to high volumes of feedback data. When it comes to serious business decision-making, small sample sizes lack statistical significance and create uncertainty. To turn qualitative data into something actionable, it needs to be quantified at a significant scale.
Imagine for a moment that your customer service team has 400 customer support conversations every day. Over 30 days, that's 12,000 conversations which are rich with customer feedback. The task of analysing 12,000 conversations by hand could take weeks—each conversation would need to be analysed and tagged with a topic and sentiment. By that time, your customers could have a different set of problems, making the analysis out of date and leaving proactive reaction on the table.
In this article, we'll address the answers to this problem and all your customer feedback analysis issues, like: Why is undertaking customer feedback analysis so important? How can you ensure the results of customer feedback analysis are actionable and unbiased? What tools and methodologies can you use to do customer feedback analysis? What use cases and case studies are there to learn from?
Feedback analysis is the process of breaking down customer data to identify customer friction points.
Business want to retain their customers to ensure they spend more, for longer. Identifying and solving friction points is essential to meeting this goal because it's essential for improving CX and tackling churn drivers.
Feedback analytics is often automated by savvy businesses. Thanks to the large amounts of NPS surveys, customer reviews and support tickets (emails and live chats in particular), most companies have a considerable volume of unstructured text data to deal with.
Machine learning-based natural language processing (NLP) can make fast work of large volumes of unstructured text. If built correctly, an NLP engine can uncover extremely granular insights from qualitative text data, and is being used by companies worldwide.
If you have a help desk software like Zendesk, NLP support ticket tagging is an essential plug in to maximise the customer feedback capabilities of the help desk.
Many companies manually analyse customer feedback data. It makes sense for some, especially if there are low volumes of data available. However, in a high volume environment, there's a whole host of problems with human-powered analytics—the main one being speed. A person like you or me can only do an analysis of so many reviews or live chats, usually meaning a potentially non-representative sample is taken.
Furthermore, a person typically has a the ability to do an analysis once a week, meaning they don't have a real-time understanding of customer complaints that can help them to proactively solve unusual source of customer issues.
Text analytics (read our complete guide to text analytics here) on the other hand, removes the need for human-powered analysis. Sentiment analysis, a subset of text analytics, categorises this text as positive, negative, or neutral and can even do 100,000s reviews in minutes. Whereas NLP can turn unstructured text into quantifiable data—for example, the frequency of reviews concerning particular topics.
This kind of customer knowledge plays a vital role in business decision-making. Allowing firms to be responsive to their customer's needs and therefore reduce customer churn and remain competitive.
Customer feedback is the secret behind the success of companies like Amazon, Apple and Google. They listen intently to customer feedback to stay on top of changing consumer needs and to continuously improve existing products to retain happy customers.
With customer feedback, you can learn what your customers like and don’t like about their experience with you. If you know your customer's pain points with your product or service, you can tweak and improve them so they better serve your customers. Continuously acting on feedback like this means your customers won't start to look elsewhere for a company to fulfil their needs, which would lose you revenue. For example, in our research into Trustpilot reviews for a number of leading companies we discovered that products arriving damaged was creating angry customers who decided to shop online elsewhere in the future.
Customer sentiment analysis is most powerful on qualitative data. Using AI, you can understand the topic and sentiment of high volumes of rich text. Find rich text from these sources:
Support conversations are the most useful data source for customer feedback. Support ticket logs (from emails, calls and live chats) contain unbiased, qualitative feedback that's an unbeatable source of customer insight. We've written extensively about why support is the most valuable source of insight, here's one such example on support ticket insight.
Reviews are a core driver of sales, as most customers look there before making a purchase. This makes understanding the reasons behind negative reviews important for business growth. One company taking advantage of their reviews is British Airways Holidays , who use sentiment analytics to make sense of their large volumes of review data so they can respond proactively to it.
NPS and CSAT are the most common ways to collect customer feedback. Understanding the drivers behind positive and negative sentiment at scale is a powerful use case for AI-powered sentiment analytics. However, NPS surveys are now considered a biased form of feedback, so you'll need to work hard to ensure your results are actionable.
Twitter, Facebook and Instagram are a minefield. Customer's love to vent on these channels and it can be damaging to your brand if not dealt with quickly. But, they're an excellent channel to collect feedback from. If you customers are already talking about you, you can collect and tag those comments yourself. If not, consider setting up a customer group to encourage your customers to discuss your product.
✉️ Monthly newsletter
🤝 Opportunities to contribute
📈 Improve customer service & CX
📚 Ebooks, articles and podcasts
👋 Unsubscribe anytime
Actionable insights get listened to and you need cross-functional buy-in to make change happen.
If our insight isn't actionable, we get this sticky situation where we think we have the purest form of customer feedback out there but no one's listening to us.
There's six key characteristics to customer insight that gets listened to: insightful, timely, granular, unbiased and not based on a tiny-weeny sample size.
To learn more, read our guide to actionable customer insight here.
In this guide, we cover manual support ticket analytics. There are four core stages:
1. Collate feedback data
2. Design your taxonomy
3. Tag every ticket (free template available)
4. Identify patterns and share your insight
We also explain our best tips for sharing insight. Sharing customer insight in accessible ways ensures they get acted upon and you can actually drive real business improvement.
Read the full guide to manual customer support ticket analysis here.
In times of intense competition in customer service, a strong focus on turning customer feedback into product and operational transformation is needed to stay ahead. Customer feedback analysis done using NLP will provide you with immediate benefits, including a deep understanding of customer issues, a clear 'to do' list for your teams to prioritise, and a measure of performance over time.
Machine learning and NLP analyse unstructured text at scale. Uncovering insights from 100% of the data set. The more data covered, the more in touch you are with your customer and the more certain decision-making can be.
Often companies make support agents tag conversations, which adds to their boredom and diminishes morale. In other companies, a customer experience manager will have to weekly reporting after an endless manual analytics task. Automation tools remove all that hard work, instantly.
It's contextualised and easy to prioritise
Know the key performance drivers and set your team to tackle the biggest impact drivers first.
Human's cannot be fully objective. Two people will read a sentence and understand it differently, which is a huge issue with manual feedback data tagging. Automation ensures uniform analytics without human bias.
With the right sentiment analytics setup, you can quickly uncover multiple levels of topics. Hierarchical topic tagging means you don't just receive information like "payment issue", but you also receive subcategories like "payment issue --> Paypal failure". Understanding at a granular level ensures you can reduce barriers in the customer journey.
The SentiSum team recently undertook a customer feedback sentiment analysis for two major industry verticals. Take a look below to see customer feedback analysis in action on reviews and you'll get a hint of what's possible when brands use this technology on customer support conversations.
There are numerous tools that can help you analyse customer feedback. Whether it's a holistic point of view of all your customer feedback channels or a specific sample of reviews that you want to tag, there's a tool out there for you.
A great place to start your search is on G2, a software review search engine. Insight Platforms also a dedicated environment for analytics software. For example, you can find a video of the SentiSum product demo on the Insight Platforms site.
For further reading, we've broken down the top 29 voice of the customer tools here.
SentiSum gives a holistic, real-time view of customer feedback channels, including customer reviews, NPS surveys and support conversations. Every data point is tagged automatically, providing insights on volume, topics and sentiment. SentiSum is a powerful method of creating customer-centric change. See pricing here.
Qualtrics is an SAP owned enterprise software that enables organisations to collect feedback at every stage of the customer journey. They enable enterprises to uncover trends and drive customer loyalty across multiple channels.
Chorus captures and analyses all customer calls, meetings, and emails to identify top performers and uncover insights that could be used as testimonials.
IBM offers many APIs for sentiment analysis based on NLP. The Watson Tone Analyzer, for example, which focuses on support tickets and satisfaction surveys and monitors agent sentiment – whether they’re polite and eager to help, and if they truly solved the customer’s issue.
Rosette has an API that uses AI to analyse natural language. They've branched out from social media to analyse entire documents, for example, the sentiment expressed by customers when they mention a specific product, company, or person.
32% of customers will now walk away from a brand they love after just one bad experience. This makes for a precarious environment for ecommerce brands. With the pandemic and other world shaking events, companies are left with an increased chance of negative sentiment at each touchpoint. In light of this, delivery, digital experience and the post-purchase brand experience have never been more important.
To stay on top of customer touchpoints, negate negative customer experience and inspire loyalty, there's three features we focus on to help you remain competitive.
Truly customer-centric change is achieved by putting actionable insights in the hands of the right team. We help you counter data silos and encourage cross-functional change by adding anyone to a daily email full of customer insight.
Knowing what impacts customers most, matters. Thanks to our sentiment score, you can see which topics are the largest contributors to negative sentiment. Accurate support ticket tagging also provides volume data, so you can quickly identify the biggest causes of complaints.
The speed of AI topic tagging enables real-time alerts. If there has been a sudden increase in a new topic, you'll know about it. Meaning you can work proactively with teams across your company to prevent the same problem happening to more customers.