Applying sentiment analysis software in customer service: how, where, and why

Every day, every week, customer messages pour into your company. It could be an account email, a response to a support ticket, a live chat message, a tweet, a product review, a voicemail, a comment on a Facebook post, a web form query… it could be any number of messages, from any number of sources.

It’s great that you’re garnering this activity. However, the volume and variety of comms that your company receives make it difficult to see the bigger picture. It’s difficult enough just to respond to each message, let alone analyse broader trends and satisfaction rates.

Enter sentiment analysis software.

What is sentiment analysis software?

There’s valuable customer sentiment behind every sentence your business receives. Sentiment analysis software makes sense of that mass of messages and the mingled moods therein.

In a nutshell, the software scrutinises inbound messages for positive or negative sentiment. It does so based on text analysis algorithms that recognise key emotive words, syntactic effect and tone.  Plus, it’s trainable, and can learn to recognise the terms that matter most to your business. (“Software bugs”, or “slow delivery”, for example.)

So, sentiment analysis software monitors your inbound messages to assess factors such as satisfaction and urgency. In turn, this allows you to evaluate customer success and better listen to the voice of the customer.

Combining analysis with automated action

Being able to analyse is all well and good. If relevant action doesn’t follow the analysis, though, there’s little point in its undertaking.

When implemented with automation, sentiment analysis software helps companies upgrade the customer experience they offer. As well as better equipping brands to understand their reputation and customer base, it also enables them to react with speed and relevance – with less strain on service teams.

As an added upshot, sentiment analysis software helps brands boost their predictive power as well as their reaction speeds, by linking customer mood and behaviour models. All it takes is sentiment and a touch of smart automation. But what does that look like?

Acting on emails

We’ll start with the most prevalent communication channel: email. Doubtless, most of your communication with customers come via email threads. Sentiment analysis software can help you deal with those customer emails more sharply, more swiftly.

Firstly, you can create different classes of sentiment: sales enquiries or technical support, for example. You can then use these classes to generate the conditional execution of relevant actions, based on keywords found and customer attitude.

So, let’s say an inbound email contains words like “interested”, “pricing”, “demo”, and “more information”. As soon as the email lands in your inbox, sentiment analysis software would recognise that this is a new sales prospect. Then, based on “if this, then that” type rules, the integrated automation would forward this message directly to a sales manager, marked as high importance.

For the business, you don’t miss messages lost in a generic sales inbox. For the prospect, their email receives the immediate attention of the employee best equipped to help them.

Turning tickets into traction

Most support teams use ticketing systems to alleviate customer queries and issues. Those tickets are useful for more than support: they’re sentiment gold mines.

Using sentiment analysis software, you can ensure urgent messages are escalated, and positive feedback presented to the right people. For example, you get a ticket complimenting your support department and containing emotive language such as “helpful”, “satisfied” or “excellent”. Sentiment analysis recognises and segments this data. Then, using rule-based workflows, the satisfied customer is sent an email asking for a product review, a case study, or a testimonial – based on what best suits your business model.

Again, this is simple stuff, but it allows you to act immediately on customer mood to generate a positive outcome. Even better, it requires minimal resource from your team.

Smart SMS systems

When something goes wrong, you want the right person to know about it, right away. In these instances, sentiment analysis software can be invaluable for damage control.

A complaint is a great example. If you receive an aggravated complaint from a customer, you want to be on the ball in rectifying the problem and preventing churn. With sentiment and automation, a detrimental, high-priority message can be detected as soon as it lands – no matter what time or channel. Based on the keywords found – “deeply unhappy”, “cancellation”, “not satisfied” – an SMS message could then be triggered to send to your customer service manager, containing extracts from the complaint and sender details.

For you, this means 24/7 awareness of business-critical issues. Your finger is on the pulse, making you better equipped to put out potential fires.

Speedy social support

Sentiment analysis software is an excellent social listening tool. First, it can monitor digital comments made about your company. Then, it can run smart automated actions off the back of what it finds.

This goes further than detecting negative or positive feedback. It might be that a customer has tweeted something like: “Anybody know a cheap alternative to [Company?] Looking to switch providers.” By looking at syntax, sentiment analysis software could identify the meaning of this tweet, and a triggered email alert to a customer service rep could then be sent.

Smart social listening doesn’t just mean that you can act on opportunities. For your brand, it’s a chance to find out how the public perceives you. You can see your company’s true standing behind a torrent of tweets and text.

Live service insights

Particularly in contact centres, it’s important to see live levels of customer satisfaction. Once more, sentiment analysis software is the ideal tool to help.

Most contact centres will be using live chat software to support customers. This ongoing influx of instant messages can be analysed line by line and session by session. The upshot of this is real-time calculations of customer happiness, chat resolutions and escalations. Plus, when combined with automation, this analysis can be used to auto-create urgent support tickets, sync sentiment ratings into customer records, and alert team leaders to assist with difficult chats.

For contact centres, sentiment analysis helps with granular analysis, as well as an evolving view of the bigger service picture.

Scratching the service

These use cases are only scratching the service of what sentiment analysis software can do. All your inbound messages can be monitored and actions applied, spanning everything from in-app reviews, to voicemails, to documents, to web forms.

The value of tapping sentiment is enormous. It’s the first step to truly hearing the voice of the customer – any time, any channel. Or, as we wrote in a previous article, “Sentiment analysis will do to big data what Google Translate did for all the world’s languages.” In doing so, it will help inform your customer experience initiatives for ongoing improvement – without the increased resource.

Useful links:

ThinkAutomation’s sentiment analysis feature

WhosOn’s sentiment analysis functionality