Agent efficiency, automation, and operational insights
Introducing the next generation of online communities. Read the announcement
Customer Experience (CX) is undeniably a crucial part of any organization’s operation. But too often, it flies under the radar. We live in an age where analytics, data, insights, and world-class CX is more accessible than ever — and that’s a good thing because customer expectations are also at an all-time high. Your customers want a seamless experience on multiple channels; if it’s not your brand giving it to them, another brand will.
In this guide, we’ll cover everything you need to know about CX analytics — what they are, how you can use them, and even how you can report on them. Let’s start at the beginning.
Customer experience (CX) analytics is the collection and analysis of customer data by businesses to improve their offerings and interactions with customers.
Customer data can include almost anything from any of your customer interactions, such as:
Phone support conversations
Social media mentions or chat logs
eCommerce goal completions
Web chat logs
Reviews on your website or other sites.
These insights will help you develop a deeper understanding of the customer journey, and ultimately improve it.
With CX analytics, you’re not simply shooting in the dark when you try to improve your offerings. You’re making data-driven choices that have a better chance of positively impacting your brand’s bottom line. Ultimately, this is what will determine the success of your CX program.
Customer Experience analytics provide a candid look at what your customers want. Customers care deeply about their experience with your brand, and they give you near-constant feedback on what you’re doing well — and where you could improve. This data can provide the information you need to enhance that experience and fix underlying issues before customers even ask you to do so. It helps you anticipate customer needs, and at the end of the day, it can seriously impact your brand’s bottom line.
There are four common ways that insights from CX analytics can impact revenue:
AHT (average handle time) reduction
Here’s an example of how one brand (name redacted) used CX data to impact their bottom line:
The brand was experiencing a higher-than-normal call volume from customers asking for credits, mainly due to incorrect invoices. Leadership believed it was a customer education issue — that is, that customers weren’t properly educated on how the first invoice was delivered.
This elevated volume caused issues: CSAT levels for first 90-day customers were lagging behind the average, and churn for first 90-day customers was out of the brand’s tolerance range.
The brand used speech analytics capabilities to identify calls and chats where customers were asking for credits within their first 90 days with the brand. This helped uncover the real root cause: not a customer education issue, but a glitch that caused autopay not to process for customers within 10 days of a billing cycle. The glitch made customers believe they had paid their balance in full when they hadn’t, resulting in misassigned late fees.
Once the brand was aware of this major glitch, IT addressed the billing rule that created the issue. This led to a direct reduction of customer churn as billing dissatisfaction inquiries were reduced and customers stopped incurring the late fees that were being charged as a result of a system glitch.
Okay, so CX data is crucial to your business. Every team, every function — indeed, every employee — probably has to think about customer experience data on a somewhat regular basis.
Despite this, many brands struggle to measure the customer experience. You're not alone if you read that last paragraph and immediately got nervous about how (or whether) you gather and interpret CX data. Accenture reports that up to 80% of enterprise brands suppose data is "unstructured." This includes customer sentiment through phone conversations, images, and other media that can't be quantifiably measured. CX data can be exceedingly challenging to process and use in this state. In other words, the issue here is that brands need to collect more data; for the most part, they manage more than enough. The problem is that they aren't sure how to draw actionable insights from their data.
If this describes the situation in your brand’s CX program, we have two pieces of good news for you. First, as we mentioned above, you’re not alone. Generating actionable insights from mountains of data is hard, and many brands struggle. But second and more importantly, there is a tried and true process for collecting, analyzing, reporting, and using CX data to improve the customer journey.
Naturally, we’ll start at the beginning of the process: setting up your CX analytics program to track the right KPIs for your brand. Simple enough. Just kidding — we know many brands have a hard time with this. But these steps will help.
The first step in gathering CX data should always be identifying your goals: fixing pain points in the customer journey, increasing conversions, or forecasting customer needs. This will give you a better idea of where to start with your existing CX data.
That’s right — existing data. Instead of investing precious resources into expensive surveys or market research and then filtering the results, focus on the CX data you’ve already collected from your countless customer interactions; there’s likely to be a lot of it. Gathering insights from these channels is far more efficient and cost-effective than collecting data through new surveys and market research groups. Furthermore, you’ll also be using data based on your actual customers, not test groups or survey respondents who can be unreliable.
Now that you’re in the right and insights-oriented mindset — start identifying all the channels and points of entry where customers interact with your brand. Both directly and indirectly. This includes:
These channels are just a few famous avenues customers use to engage with brands. Your company may have more, fewer, or different channels. But regardless of which particular channels your brand uses, you should focus on aggregating CX data from as many as possible to create a holistic view of the customer. A comprehensive view of a single channel won’t give you the whole picture; you need to think about every channel.
Phone, chat, and social media interactions are extremely rich in data, accessible, and popular. But that doesn’t mean you should ignore your other channels. Every channel serves a specific purpose. Furthermore, each also presents a different version of the customer journey, and you’ll need to gather insights from all of them to take your CX to the next level. Tapping into customer journey analytics is your best friend here.
You may find that your most important customers aren’t always the loudest, and they might interact with you through an obscure channel you never thought about.
Okay, so it’s important to gather actionable data across every customer engagement channel — but how do you actually collect and organize this CX data?
You could go channel by channel and manually pull insights, but this is inefficient, time-consuming, and requires extensive knowledge of each channel. Even once you collect data, it will be hard to analyze because unstructured data varies widely across channels. For example, the CX data you pull from support chat won’t have the same format as the info you draw from your online community — which will differ from data gathered from your social media interactions.
It may be beneficial to use an omnichannel solution to compile and organize customer experience data. Analytics for customer experience may be hard to interpret if you’re looking at several data types, so having it all in a unified format on a single platform can make analysis easier.
Got all your data together in one place? All right! Look at you go! Now it’s time to do some analysis to determine what’s important and what’s not. Don’t panic — we’ve got you covered with some easy steps.
Looking at all your data in one place for the first time can be overwhelming. There’s a mountain of it; where should you start? The best way to get started is to prioritize whichever areas or channels receive the most interactions. Social media? Web chat? Phone? Email? SMS? Something else? Whether it’s one on the list above or not, that’s the place where most customers are going to interact with you, so that’s where you should start.
But remember: just because you start in one place doesn’t mean you get to ignore the rest. Once you have a handle on one channel, examine the others as well; there will always be more insights as you look deeper into your CX analytics.
Based on these interactions, identify at a high level what causes the most friction in the customer journey. Start with obvious insights, then move to more hidden ones. One excellent method here is to break your CX data out by cohort. Look for customer trends by channel, demographics, frequency of interaction, survey and review scores, and anything else you can think of. The objective here is to identify as many friction points as possible and determine which customers are having the most challenging time with the broken customer experience.
If you aren’t getting as much as you thought out of the previous step, you may need to give your data some context and enrich the story you’re trying to tell. That’s where metadata, or secondary data, comes in. This includes any information about customer interactions you didn’t look at in your primary analysis. For example:
Clickstream data (e.g., web visits or engagement)
Processing this metadata, even in an unstructured form, is valuable because it provides insights without you having to solicit feedback. But even more importantly, it allows you to extract information about the conversation — keywords, topic, sentiment, and more — from the conversation itself. Not only is this helpful in surfacing trends via commonly used keywords, but it also helps organize the conversations to build greater understanding and context.
For example, seeing that conversations mentioning "return" and "product name" are spiking allows you to notify the appropriate teams, perhaps product and sales, that certain products are having quality issues. Then you can take affected products off the shelves, create a script for customer support agents, and thus improve efficiency.
By now, you’ve probably identified several areas of improvement in your brand’s customer experience. If you haven’t already, determine the root cause of each point of friction and gather context to help you determine the severity. By understanding the source and scale of the issue, you can get a better idea of what it will take to fix it, as well as the impact of that resolution.
Make sure to prioritize friction points based on the severity of the issue, how easy it is to fix the problem, and the expected impact of the resolution. Don’t try to do it all at once, especially if you’ve identified a number of issues. That won’t work well, and you’ll risk making things even worse. Instead, tackle the low hanging fruit first before moving on to bigger issues.
Having implemented a solution for at least one issue, look back at the analytics that you used to identify that issue. Has the situation improved? Can you point to a clear reason why or why not?
Creating and sustaining a CX program requires buy-in from key stakeholders across the organization. In order to get that buy-in, you’ll need to show clear, quantifiable results (or at least demonstrate that there’s a quantifiable issue to be solved). As much as possible, make sure to tie your data back to business outcomes. To do this, you’ll need to make clear, concise reports for your CX program.
Quantifying impact will hopefully yield improvements in KPIs. Here are a few of the KPIs your brand might care about (although every brand is different):
Average handle time (AHT): It’s important to review high-volume friction points to see how average handle time at those points differs from the overall average. Friction causes handle times to greatly increase as agents struggle to provide answers. Root causes for increased handle times include agents needing supervisors, longer-than-normal hold times, and higher-than-normal transfers. For example, customer inquiries about billing confusion have handle times about four times the average.
Customer satisfaction (CSAT): There is a strong correlation between low CSAT scores and high-friction journeys. For instance, about 80% of customers who call about auto-renewal have a CSAT score of zero.
Call deflection rate: When reviewing friction points, you should always ask whether the customer can self-serve to solve the problem. Can the business improve education on common friction points or eliminate friction altogether? One example is customers calling about login issues; it’s relatively easy to either give them a way to self-serve or (if the complaint persists) make the login process easier. That way, customers don’t have to call or chat for assistance.
Customer effort score (CES): There is also a strong correlation between high customer effort scores and high-friction journeys. For example, customers who call about billing issues have customer effort scores of “strongly dissatisfied” 60% of the time.
Net promoter score (NPS): Low NPS scores also show a strong correlation with high-friction journeys. For instance, 80% of customers who call about auto-renewal have NPS scores of zero.
Sentiment: Sentiment is an emerging KPI that represents where friction is happening in the experience. Sentiment analysis is performed on each interaction based on the words spoken during the interaction (recognized based on the transcript). Sentiments are weighted more heavily at the end of the conversation.
Delivering CX insights is a critical step of a CX program. Every company has a different preference for how to deliver data and insights. Whether it’s via excel sheet or a polished agency presentation, it’s crucial to organize your work and tell a cohesive story.
In order to ensure company-wide adoption of the CX program, aim to make the report concise and easy to understand by categorizing and limiting the number of insights you share. Stakeholders don’t want to read long diatribes about customer experience; they want short, clear, actionable insights — after all, that’s the very purpose of CX analytics.
One successful reporting method is to group the insights by a subject or theme, such as business function, experience, product, service, season, membership, or customer demographic. Only include the six to eight key insights that will be the most impactful and garner the most attention. Including too many in a single report can give executives and employees insight fatigue.
While these insights provide opportunities to drive immediate impact and revenue, some teams may be reluctant to accept your findings at first. In these cases, it's best to provide teams with insights they are already familiar with or may even already know. This will get them to buy-in on what you are uncovering. Once you’ve established a rapport, you can slowly start sharing more dramatic insights that may not align with their business plan.
Title: What is the focus of this report? Provide a high-level theme or subject.
Statement of insights: List the insights, current status, teams impacted, KPI measurement, and priority level.
Customer verbatims: Give two to three customer verbatims with each insight. Remember, less is more when it comes to impactful insights!
Details and human elements: Provide some analysis for each insight, identifying how friction affects KPIs (handle time, transfers, holds, etc.), as well as the actions you plan to take. Also, remember that human elements like quotes and anecdotes make your report more readable and help build understanding and relationships — keys for a cross functional team like CX.
Wins: If there are any customer experience improvements that have been executed from previous reports, celebrate them! This is where you give the proper business functions credit for their work.
If you understand the components of a CX report but want a visual reference of what one may look like in a real-world application, here’s an example of a CX report discussing a brand’s membership program:
The first part of the report is a simple table that conveys friction points, action statuses, geographic areas affected, how results are measured, and the priority of addressing the friction points.
The following slide is broken into three sections. On the left are some sample customer comments to contextualize friction points. A bar chart in the middle displays how consumers are asking to join, with some additional information for reference. Finally, the right side of the slide shows actionable steps to resolve the friction points.
The final slide of this CX report is similar to the first. It shows the initial friction point, the action status, and the final results. This example of a CX report is focused on a single area of a business. Still, messages can contain several sections like this to cover additional products, services, or other interaction points.
Once you’ve created a CX report, don’t be afraid to distribute it across your organization for at least three reasons:
Every employee plays an integral role in the customers’ experience, so they need to identify and understand the voice of their customers.
They may have additional information regarding insights you brought up, and you can use their input when implementing changes to improve the customer journey.
Employees are more likely to buy into your CX program if they can see the impacts of their efforts.
The most important part of any CX program are the outcomes. A results-driven CX program helps improve the entire customer experience while also impacting the company’s bottom line. Here are six steps almost any brand can take to efficiently and effectively improve their customer journey.
As we said above, it’s important not to try to take on every issue simultaneously. If you have multiple pain points in your customer journey, prioritize the low-hanging fruit and the solutions that will immediately impact revenue. Implementing the solutions that directly impact your brand’s bottom line — perhaps by streamlining the customer care process to make it more efficient — will give your CX program the credibility and buy-in it needs to tackle more complex problems.
Your customers and customers are speaking to you all the time. Listen. Do they strongly prefer to interact with you on a particular channel? Prioritize that channel. Do they select a specific button design? Use that design. Are they primarily interested in a particular resource? Promote that resource.
Tracking consumer data will help keep customers happy because it will enable you to implement changes they want without asking for them. That’s the power of using CX analytics.
Tracking general consumer preferences is one thing, but getting specific with customer segments is another. You’re probably already tracking demographic data and personas; you should use that information to personalize the customer experience. If they’re logged into your site, greet them by their name. Make landing pages, ads, and content specific to your target group. If you know they have a particular concern that might not be as important to other groups, talk about it in your content.
The more customers feel your brand cares about their specific interests and concerns, the better your CX will become.
A great CX program is proactive and reactive, and part of that proactivity is listening to what your customers say even outside of brand interactions. One way to do this is to combine CX analytics with social listening. With a social listening solution, you can identify trends in conversations across social media platforms whenever they touch on a particular topic. In other words, social listening provides the same insights on social media that CX analytics provides in customer interactions. You could listen for chatter about your brand, industry, product, or even competition. All this information is crucial to predicting what customers want and how you can best give it to them.
Follow up with your customers to make sure they’re still enjoying their experience — especially if it's an ongoing relationship. At this point in the customer journey, it’s not a bad idea to solicit their feedback through a survey, although you should never let surveys replace omnichannel analytics. Of course, make sure you’re not annoying customers by following up too often. You can play around to strike the right cadence to stay relevant without being overbearing.
The real pros in customer experience don’t just track their brand’s CX, but those of their competitors. Now, you’ll want to do this only once you’ve already developed a healthy, functioning CX program, as it can take a lot of time and energy to accomplish. But once you do, tracking the competition can yield further insights you otherwise wouldn’t have gotten.
Perhaps competitors are doing something that never occurred to you, and their customers love it. That’s an excellent opportunity to improve on your end. Maybe they’re doing something wrong and don’t realize it. You can make sure not to make the same mistakes. Perhaps they’re implementing a change to see how it plays out. You can see what customers say without having to test it yourself.
Regardless of the insights you uncover, staying on top of the competition is an excellent way to make sure your own CX is better than theirs.
Collecting valuable CX data requires tools that provide insights across several channels, so you can get a full view of customer experiences regardless of if they’re interacting with you through social media, phone support, a community forum, or chat on your website. Khoros offers several solutions to help you manage these customer touchpoints and compile the information into a single dashboard so you can view and analyze CX data.