Advanced AI & NLP

The potential for machine learning to elevate the customer experience has everyone buzzing. AI-powered text and sentiment analysis can be an incredible solution for specific problems that CX pros face. 

But how do you know when the time is right to move to the next level of CX? Are there new tools you can purchase to step your game up? How do you know they’ll be worth it? 

There are clear signs that your CX program is ready for, and your company could quickly benefit from, text and sentiment analysis. And we’ll delve into them here.

Before we get going, some definitions:

• Text analysis takes qualitative customer comments and determines relevant themes. Software companies might see themes such as ‘feature request’, ‘bug’, or ‘pricing’. This allows you to quickly see what your customers are focusing on, and then dive in to see what they’re specifically saying about each topic.
• Sentiment analysis offers micro and macro insights into how your customers are feeling about your company and products. It determines whether the text received for each text theme is positive, negative, or neutral. It also analyzes the comment as a whole, assigning sentiment to the entire verbatim text.

Let’s look at the 7 signs text and sentiment analytics will be worth the investment for your company. 

1. You have a mature or quickly-maturing CX program.

Those of you considering text and sentiment analytics probably already have a few key elements in place:

•  A customer experience strategy and a Voice of Customer listening system
• A C-suite sponsor who has been fostering a customer-centric culture across the whole company with NPS as the guiding star
• A system asking for feedback through the entire customer journey 

Now that you have a relatively mature CX program, you’re wondering how to extract even more value out of it.

2. You receive 500+ comments per month (or you’re headed there.)

Ideally, you want to listen to all of your customers – not just a sample or the first to respond. In reality, at a certain point the sheer volume of incoming customer feedback is more than a CX program can handle without an upgrade. You know this is the case when:

1. You feel excitement and dread regarding the amount of feedback you receive.
2. You’re anticipating a whole lot more comments soon.
3. You’ve even had to cap the number of comments you receive in a day to avoid being overwhelmed with the task of organizing and responding to everyone.

Overwhelming amounts of feedback is an amazing problem to have, but a problem nonetheless. Using text and sentiment analytics, you can turn unstructured qualitative feedback, like NPS comments, into organized insight in a matter of minutes.  

Text and sentiment analytics allow you to analyze customer feedback using Natural Language Processing, looking something like this:

By combining text and sentiment analytics, you can search negative comments and quickly assess, for example, that 80% of your negative comments are about pricing. Or 45% of your customers in the Northeast region are talking about slow delivery times. That summary lets you know where to focus resources, and how quickly you need to make the change relative to other company priorities.

3. You’re sitting on a goldmine of feedback, but unable to get actionable insights.

Do you have a backlog of comments waiting to be read and sorted? Or maybe you’ve skimmed a few comments to answer the urgent ones, but you keep putting off the others.

One of our clients came to us with NPS survey comments from thousands of users. But rather than mining that information, they were running focus groups to prioritize feature requests because it was easier. They were duplicating efforts to get information they already had but couldn’t access and act on.

“The two biggest mistakes [in CX] are not doing qualitative research in the first place and then not putting it to use.” –Morgan Brown, Product Manager at Facebook and coauthor of ‘Hacking Growth’

If you’re feeling this pain, it’s time to automatically mine the insight from that pile of comments you’ve been sitting on. Turn anecdotes and hunches that you’ve got about your customer experience into evidence-backed insight by using. And do it quickly with text and sentiment analytics.

Source: CXInsight™ Dashboard

Sliced and diced organized feedback is easily available with many platforms that offer text and sentiment analytics. Doing this can help you understand the root cause of trends – like the needs of different customer personas or geographic regions – more comprehensively.

4. Manual feedback organization & categorization is insightful, but painfully slow.

While some customers duplicate efforts between data gathering and focus groups to get insight, other CX pros just bite the bullet and spend hours reading customer comments, labeling them, and funneling them into an unwieldy spreadsheet. They’re understandably frustrated by how difficult it is to get actionable insight.

By using text and sentiment analytics, humans can get huge quantities of customer feedback sorted and analyzed at the push of a button. Better yet, computers don’t have bad days or lose focus.

Once organized with tags, your time is freed up to look at the themes and trends that arise from the noise, then create actionable strategies based on those insights.  

Now you can jump straight into action and the interns can work on more interesting, valuable projects!

5. Your CX program lacks a real-time issue detection system.

An important element to providing a good customer experience is making sure any issues are handled quickly and efficiently. If you can detect and address them before your customer has a real issue, your CX program has paid for itself.

One of the benefits of having text and sentiment analysis is that your data and insights are updated in real-time. This means you have a new issue detection system.

Source: CXInsight™ Dashboard

This works best for a more mature customer feedback program with an established baseline, or status quo. For example, you know that on any given day, in any given geographic region, about 10% of your comments are tagged with ‘out of stock’ as an issue. When you check in and see that in Texas, 25% of comments coming in are tagged ‘out of stock’, that raises a red flag. You can immediately dig into specifics, read through the verbatims, and send those comments to the right people for follow up before the issue blows out of proportion.

The CX dream of being proactive in solving issues can be achieved with the help of automated organization of qualitative feedback.

6. Your internal teams aren’t agreeing on CX priorities.

It’s a given that successful companies focus on customer needs and experiences. The question is: is everyone at your company seeing the same information in the same way? If not, you’re wasting time and costly resources with competing priorities, and it is definitely time to invest in tools to fix it.

By having your CX tech parse the text and sentiment of your 1K+ daily inputs of customer feedback, you can democratize the information and insights across every team at your company. And that will ensure team leaders can quickly align to address the right priorities. So product development and customer support will be on the same page, and features will get developed (or possibly de-bugged) to meet the most important needs of the customer.

How does that happen? Feedback from every customer touchpoint is analyzed, from in-product surveys to emails. In this example, support ticket subject lines are auto-categorized and everyone from support to service to product to the c-level can see what issues are hot items to address.

Source: CXInsight™ Dashboard

Looking at the text analytics, it quickly becomes apparent that 15% of the support tickets are related to bugs that need to be addressed. On the proactive front, product could also delve into comments tagged “feature request” and focus on user concerns about UX/UI.

7. You need to demonstrate the ROI of your CX program.

Companies are eager to hop on the CX bandwagon, but it can still be a fight to get the proper resources to make a CX program thrive. You’ve probably already shown the C-suite the correlation between CX and revenue growth, but there’s pressure to squeeze a little more ROI out of what you’ve established. 

Investing in a tool that pulls ROI from data is an expense. But it’s a more strategic spend than, and offers more immediate follow-up and action, than  performing passive data review and organization. It’s also a moredirect value-add and much less expensive than hiring a third party human operation. 

The cascading effects throughout the organization will increase ROI in the long-term as well.

• Product teams can prioritize and build with evidence-based confidence. 
• Marketing teams will gain an understanding of different personas and see customers excited to spread the word about your business. 
• Support and operations teams will have early warning of potential issues and have more context when dealing with problems.

In the end, qualitative data is crucial to extracting value out of CX initiatives. Having more data from engaged customers should not be an obstacle. 

Is this the point?

Are you seeing any of these 7 signs when you look at your company’s CX program? If so, do a cost benefit analysis. Typically, once your program has matured, the cost of tools that create actionable insights out of customer feedback are far cheaper than the cost of misaligned resources and long delivery times. Text and sentiment analytics make the resources you put into CX initiatives efficient, and turn the large quantity of unstructured data into an advantage by mining insight that would otherwise sit in limbo. Move this tipping point in your favor.

Wootric (now InMoment) uses CX metrics—Net Promoter Score, Customer Satisfaction and Customer Effort Score—to monitor customer experience for high-growth companies. We take a customer-centric approach to survey design. For example, our modern 2-question Net Promoter Score survey invites customers to elaborate freely on the reason for their score. We deliver millions of surveys that achieve response rates of 30-40%, generating thousands of pieces of unstructured customer survey feedback each week.

Why Is Survey Feedback Important?

Because when you communicate directly with your customers, they can identify exactly what works, what doesn’t work, and where the pain points are that may be detracting from their experience. Honest feedback gives you the insights you need to make improved business decisions and optimize the customer experience. As such, the right customer survey can play a significant role in increasing customer retention and helping your organization reach its goals.

Customer feedback comments are a treasure trove of information that can help a company shape their product and service for success. Until now it has been difficult for a Customer Insights Manager or customer experience management (CXM) teams to mine and aggregate qualitative data for insights that can guide business decisions.  

Auto-Tagging with Sentiment Analysis

We recently announced early access to a new product feature: auto-tagging. For auto-tagging, we use our homegrown machine learning system along with Google Cloud Natural Language API to automatically categorize open-ended customer-survey feedback that our customers get as part of their NPS, CSAT and CES programs. The goal is to help companies put some structure to all of this qualitative data. We have a long list of customers eagerly waiting to get their hands on this feature. It’s a good problem to have.

In addition, we are developing the ability to identify the sentiment of the feedback. The goal is to determine not only what the customer was talking about, but to say whether the feedback is positive, neutral or negative. It is particularly complex to decipher multiple “sentiments” within a single comment.  

Here is an example feedback comment that we received in response to a Net Promoter Score survey on our own production application (we practice what we preach):

“Setup guide for customizing social sharing on iOS SDK was confusing. Diego reached out with sample code which helped a lot.”

Wootric (now InMoment) is a SaaS product, so our auto-tagger uses a SaaS data training model and applies three tags to this survey response (Documentation, SDK, People), and assigns a NEUTRAL sentiment for the feedback as a whole. This obviously is pretty good, but we want to do more.

Wouldn’t it be nice if we could dig deeper into survey feedback and apply sentiment for each tag as well? In the above example, the customer was not happy with the SDK Set-up Guide, but was pleased with Diego’s assistance.  This nuance is buried under the overall NEUTRAL sentiment. Ideally, the Documentation and SDK tags would be identified as having negative sentiment, while the People tag would be positive.  

We Can Identify Sentiment Associated with People, Team, Organization or Location

Buried survey feedback is not a trivial problem to solve. However, using Google Cloud Natural Language API’s latest feature called “entity sentiment analysis” we have made progress. We can already get sentiment for entities referenced in feedback where an entity is defined as People, Team, Organization and Location. In this case, Diego is an entity of type People and positive sentiment is correctly attached to it.

Example of Auto-tagging an InMoment NPS Survey Response

A Business Use Case

Our customers often trigger a CSAT survey using our incoming webhooks and workflows when a support case is closed in their CRM system like Salesforce or Zendesk.

We notice that survey-feedback responses often reference a team or specific person that the customer has engaged with. Auto-tagging this feedback as “People” with applicable sentiment will provide these companies with an easy way to measure and track how customers are feeling about the people aspect of a company’s Customer Success or Support program.

The challenge of open-ended feedback

Qualitative feedback in survey responses: Marketing, Product, Customer Insights, and Customer Success teams love it! There is nothing quite like hearing authentic, open-ended comments about your product or service directly from customers in their own words. Nothing is more powerful than hearing from the customer first hand: It drives action.

Individual anecdotes tell a story that can provide color and context to business metrics like Net Promoter Score, but how do you make it actionable? How do you aggregate qualitative data to see trends and get insights that can drive business decisions?  To a certain extent, this has always been an issue for voice of the customer feedback programs. However, two broad trends are driving an increase in qualitative data and creating more urgency. As a result, the problem of “metricizing” open-ended feedback is now more acute.

Customer experience survey trends that are driving the need for NLP

First trend: The shift to customer-centric surveys.

It has become more and more difficult to persuade customers to respond to traditional company-centric surveys — the multi-question monstrosities that ask customers to rate attribute after attribute on a 5 or 7 point scale.  Long, boring, tedious — and frustrating. Response rates in the single digits are common.

I recently visited the website of a major department store and was prompted to fill out a pop-up survey with over 30 (!) questions. I thought I’d get an opportunity to tell the retailer what was important to me — how much I loved their shoe selection and that I’d had a disappointing experience in one of their stores. I didn’t finish the survey.

In an effort to improve response rates, many companies are now thinking about the survey experience from the customer’s perspective. A Net Promoter Score survey that asks one question and lets a customer provide open-ended feedback is a better user experience — and customers are more likely to respond.

Wootric is modern customer feedback management software that allows businesses to gauge and quantify customer loyalty through proven feedback metrics such as Net Promoter Score (NPS), Customer Satisfaction (CSAT) and Customer Effort Score (CES). We are firm believers in the customer centric approach.

For example, here’s an NPS survey that Wootric presents in-app (we also support mobile, email and SMS) that usually takes a user less than 30 seconds to complete.

Second trend: Hearing from as many customers as possible.

Traditionally, customer research efforts were satisfied with feedback from a statistically significant sample of customers. Now that any customer has the potential to influence the trajectory of a business — whether taking their complaints public on Twitter or writing a glowing review on Yelp or G2Crowd — more companies are proactively asking all customers for feedback. This instantly opens a direct communication channel, and gives companies the opportunity to build, monitor and leverage  relationships with any and every customer.

These trends put the onus on companies to make sense of a firehose of open-ended feedback, and that is tough to do.  Dedicating resources to tagging and sorting hundreds, even thousands, of comments is expensive and just doesn’t scale.

Natural Language Processing to the rescue

Natural Language Processing (NLP) is a type of machine learning that enables computers to understand human language. You can read how Wootric applies NLP to customer feedback like NPS and CSAT survey responses in this article.  And here are three familiar examples of NLP at work:

• Machine translation like Google Translate.
• Sentiment analysis — sifting through all those Twitter posts to analyze how people feel about the latest iPhone, for example.
• Chatbots — the customer support “agents” that have become the first line of interaction when you reach out for tech support online.

Unlocking the power of open-ended feedback

NLP is solving the unique challenges in the field of customer feedback management using text and sentiment analysis. Being on the forefront of this innovation means Wootric customers are seeing those benefits now. We work to free our customers from the time and expense required to manage this data. We use text and sentiment analysis to surface and aggregate insights for our customers, helping them to prioritize resources and route responses for follow up action. Read more about what we are up to here on the Google Cloud Platform blog.