CX Analytics


Best-in-Class CX Programs Go Beyond Surveys & NPS Scores

I am an omnichannel conversational data enthusiast (someone who believes it is crucial to listen to your customers on all communication channels). I find it difficult to hear most companies over-rely on surveys and NPS scores to understand the customer experience. This is one of the many voices of the customer. I am not saying surveys and NPS scores are a bad thing, but it is just the start of understanding who your customer really is.

My passion is teaching companies how to listen to their customers on all channels and how to turn insights into greater returns. I like to think of myself as an evangelist as I teach companies how to develop a holistic approach to listening to customers and recognizing missed opportunities to hear from customers that do not participate in surveys. On average, the customer response rates for surveys is around 10%-15%. This low response rate begs the question, what about the unsolicited feedback customers are leaving on Facebook, Reddit, Instagram, Twitter, Glassdoor, Yelp, Google Reviews, Apple Reviews, Text Messages, Emails, Phone? Companies often curate surveys to post on a few of these channels but still leave the unsolicited data behind to sit in silence. Why should you care about hearing everything the customer has to say? Here’s why.


Karen goes to Good’s Coffee. Karen is a loyal customer who comes to Good’s Coffee every day. She’s always in a rush to get her coffee and get to work. Karen orders an espresso with a muffin. Not only did Good’s Coffee delay Karen’s order 20 minutes, but they also gave her the wrong order. What do you think Karen does to voice her frustration?

A) Fills out the Good’s Coffee email survey or social media survey.

B) Posts on any one of her social media accounts and tags Good’s Coffee.

C) Calls Good’s Coffee customer support.

D) Sends an email to Good’s Coffee customer support.

…she can do all 4 and much more.

If Good’s Coffee is able to hear Karen on all platforms, they can easily identify this friction point of baristas getting orders wrong. There could potentially be an issue with POS systems or lack of training causing baristas not to pay close attention to orders. The point of this example is simple: it is up to your company to aggregate and normalize all solicited and unsolicited data. This gives your company a full picture of what your customers are saying about your products and services so you can make impactful CX improvements that actually matter to your customers. I understand this is easier said than done. Making impactful CX improvements based on your company’s data can be a big undertaking and one that companies have not had the ability to fulfill. I know how daunting this first step, gathering the data, can be without the proper system and tools in place.

Achieving an omnichannel view of all customer feedback is only as good as a company’s Customer Experience Analytics solution (CX solutions). CX solutions for most companies range from an organization’s key performance indicators to NPS scores. At Topbox, we push the boundaries on how we help organizations like Aetna and Western Union listen to customers. Using AI and machine learning, we help companies gather, aggregate, and normalize conversational data from any platform customers use to voice their opinions of products and services. We believe it is important to listen to your customers on an omnichannel level. This includes solicited (surveys/NPS scores) conversational data and unsolicited (social media platforms, reviews, email, phone, chat) conversational data. As interested as I am in companies that embrace new CX solutions like Topbox, I find companies that do not embrace them even more interesting.

Whether you already love CX solutions or feel energized to learn how to make consumer data manageable, Topbox would love to partner with you. We can help your company capture all solicited and unsolicited data, consult you on how to build workflow and reporting systems, but most importantly help you to better serve your customer and generate revenue in the process. Request a Topbox demo with a real live person.

VoC data

Moving the CX Needle – Why UNsolicited VoC data is So Important

For more than a decade companies have relied on surveys to be the primary, and in many cases the only, method of listening to customers.

This solicited form of data hatched an entire industry with companies like Qualtrics and Medallia dominating the market. There have always been inherent flaws in the survey process: low takes rates, sparse verbatim comments, and polarity bias are just a few. However, having quantifiable customer feedback in the form of NPS, CSAT, or other surveys was the only semi-reliable source of data for VoC programs.

Surveys serve a purpose, and they won’t likely be discarded anytime soon. That said, companies are waking up to the fact that they are sitting on a mountain of unsolicited feedback in the form of phone calls, chat sessions, emails, app comments, SMS, social media, and other channels. The combination of solicited and unsolicited data can be a game-changer for companies that are paying attention.

Those conversations that happen across all the unsolicited channels occur organically in the normal course of a relationship between a customer and a company: nobody has to ask a customer to pick up the phone or jump online to communicate with a company. There’s a reason for these interactions – customers have problems!

So while many companies are still trying to understand how customers feel (at least that diminishing number of customers who respond to surveys), there is actionable data that can help identify the origin of customer issues. Pinpointing and fixing the source of customer issues is how a company can affect how a customer feels.  Measuring is one thing, managing is totally different.

There have been three primary challenges for VoC teams trying to get their collective arms around this immense set of unsolicited data:

The first is access. Nearly every company has an array of communication platforms that sit in different business units (silos).  Turf wars, data hoarding, and politics can all come into play when a VoC team attempts to gain access to these precious resources.

The second challenge is disparate formats. Each of those communication platforms has its own database. Normalizing up to a dozen data formats is a daunting task for even the most advanced data teams. The third challenge is classification. Organic feedback generally comes in the form of unstructured, conversational data. Inconsistent classification across those communication channels leads to apples and oranges in contextual topic generation.

Enter the “agnostic aggregator.”  New technology like Topbox gives companies the ability to leverage all that unsolicited data into actionable insights that actually improve the customer experience. A pure-play aggregation and analysis platform like Topbox sits on top of a company’s entire customer communication ecosystem, ingesting every conversation, normalizing the data, and classifying consistently with topics that are relevant to that business.

This delivers consistent, omnichannel sentiment analysis, friction point identification, anomaly detection, and trend analysis – all in an omnichannel environment.  Agnostic aggregation of unsolicited, conversational feedback is a true enabler, unleashing the full potential of VoC teams.

Companies will continue to use solicited data from surveys, but businesses that want to move the needle must start looking to new technology to tap into the insight-rich conversations their customers are having every day.


Bounced Customers are Unhappy Customers

At Topbox, a big part of our analytics is helping our clients understand when their customers are victims of “Bounce”.  The definition of Bounce in this context is important.  We’re all familiar with the concept of transfers in a contact center environment where customers are transferred to and from departments such as customer service, tech support, account services, retention, etc.   Think of “Bounce” as a macro level transfer, where customers are interacting with a company outside of the contact center, but are then forced to work through the contact center to resolve an issue that stemmed from the original interaction.

Let’s walk through an example to help clarify what we mean by Bounce. A customer walks into a local bank branch to open a new account.  The teller goes through the account set up, takes the initial deposit and tells the customer that a debit card will be sent to the home address within 2-3 business days.  The 3rd day passes and the debit card hasn’t arrived.  The new customer finds the 800 number on the bank’s website and calls customer service. The agent in the contact center finds the customer’s information but notifies the customer that the company’s policy for shipping cards is 5-7 business days.

You can imagine what the customer’s immediate perception of this bank may be.  She had a very disconnected experience between the branch and the contact center, and worse, she has to wait another 3-4 days to get her debit card.  “Bounce” is the result of a mistake made at the branch that created a call to the contact center.  We call this Retail Bounce.  You could substitute any brick and mortar store for the bank branch and think of myriad examples of Bounce across retail environments.

Another example is Digital Bounce.  A customer goes online to purchase a new pair of shoes. After determining the model and size needed, the customer attempts to add the shoes to his cart but cannot.  After attempting to add the shoes to the cart several more times, the customer decided to chat with the contact center to finish the transaction.  The agent completes the purchase and the customer is ensured he will get his new shoes.  Everyone’s happy, right?  Think again. There is a ton of research around the number of customers who abandon an online purchase and never return to complete the transaction.  In fact, some of this research shows that nearly 30 potential buyers never return or find another way to complete the purchase (like the contact center)  for every 1 who does.  That’s a boatload of lost revenue.

The net of all this is that Bounce is costly in many ways. Lost revenue, added cost-to-serve, poor CX and the resulting customer perception all have a negative impact on a company’s bottom line.  Understanding and resolving the points of failure that lead to Bounce can significantly improve a company’s performance.  With Topbox, we leverage NLP and associated metadata to highlight instances of Bounce with the necessary granularity to identify the root cause.  The elimination or even reduction of Bounce in your business will result in lower cost-to-serve and massive improvement for Customer Experience.


One size does not fit all: the 3 building blocks of a hyper-personalized CX

Personalization: The holy grail of customer experience, and one of the most difficult tactics to master. It has become something every company attempts to achieve due to its value to customer loyalty. According to a 2016 study by Accenture, 75% of consumers are more likely to buy from a company that recognizes them by name, recommends options based on past purchases, or knows their purchase history.

A personal experience creates for a more pleasant customer experience, one that helps customers feel unique and appreciated. These experiences can also provide a more seamless experience which allows a customer to achieve their intentions quicker, wasting less of their time.

The definition of personalize is “design or produce (something); to meet someone’s individual requirements.” Keyword is individual. I often see companies’ market to cohorts of customers thinking they are providing personalization, but this is incorrect. In order to be personal, a company must know what makes each individual customer unique.

The best way to learn what makes each customer unique is to have a conversation with them and listen closely. Conversing is how people learn about one another and build a relationship; companies should be no different.

In the past, previous purchases or online engagement is what companies have used to base a customer’s individuality on. Today, with new Natural Language Processing (NLP) technologies like Topbox, companies can have these conversations and gather insights about every customer across any channel. Learning this information used to be the most challenging step. Today, that is no longer the case. Leveraging the information to create a personalized customer experience is now what’s difficult.

In order to achieve personalization, you must understand the components that formulate the tactics and where your company stands within each.

Below I cover the following three components and their relationship to personalization: Marketing, Channel Experience, and Relationship Level.

Marketing – How targeted is the personalization?

Mass Marketing – Not personalized to any cohort or individual consumer. These forms of marketing often have the greatest reach but the lowest impact. The focus here is quantity over quality (i.e. mass media, billboards, etc.).

Cohort Marketing – Tailored to a specific cohort or group of customers that have been identified based on a unique, shared attribute. This is the first step towards personalization and may feel personal to some customers (i.e. unique ads based on demographics such as location or income, promotions for customers who buy specific products, etc.).

1-to-1 Marketing – True personalization; marketing at the most granular level. In order to achieve this level of personalization, a company must know unique information about a specific customer. After all, individualized experiences can only be done by understanding who the individual is. Something to keep in mind when deploying these experiences: there’s a fine line between being personable and creepy, and that line needs to be established for every customer in the beginning.

Some examples of good 1-to-1 marketing would include: offering a product or service based on the individual’s current lifestyle or preferences, greeting them by name, or going as far as asking them how they felt about a previous product or service and why they chose you.

Channel Experience – How fluid and consistent is the personalization across channels?

Disparate – Each channel is a different experience which provides disjointed personalization. As a customer provides more information within one channel, there is no passing of information to another. This creates a frustrating and fragmented experience for the consumer.  

Inconsistent – Having connected, personalized experiences across some channels. Most commonly, digital channels will have consistent experiences from one digital channel to the next (web, app), but when outside those experiences’ personalization is lost. This is a step in the right direction, but a sign that not all business functions are aligned or sharing data.

Omni – When personalization is consistent across all customer touch points. This means there is information shared across multiple business functions within a company to ensure the customer receives the most premium and personal experience.

Relationship Level – What information is used to create a personalized experience?

 Simple Personalization – There are varying degrees of personalization. Simple personalization is when basic customer information is used to create a personalized experience. This is information like name, location, and any other general information that is usually public. Since this information is unique to the individual this is still considered personalization, however, the information didn’t require a relationship to acquire.

 Complex Personalization – Unlike simple personalization, complex personalization is only possible with information that is derived from interacting with a customer. These are things like preferences, lifestyle, and household status (married, family, dog, etc.). This is where it’s pivotal that you have developed a solid relationship with a customer and are not overstepping your bounds with personalization. When done properly, this gives a customer the feeling a company is truly appreciative of their business.

There are many other aspects of personalization that companies may have to formulate, but at a high level, these components are relative to most. Because personalization is constantly evolving and changing, there is always a moving target – automating personalization as much as possible will help you create individualized experiences with precision and speed.

Personalization is complex due to the uniqueness of every consumer. This is what makes it so interesting and challenging. Luckily, with the power of new technologies, companies can achieve true personalization.

Energize Your CX Program with New Methods for Old School Root Cause Analysis

I think most would agree that the overarching purpose of any Customer Experience (CX) or Voice of the Customer (VoC) program is the identification and resolution of points in the customer journey that have a negative impact on brand perception. When a customer has a bad experience, there’s almost always an underlying reason that can be pinpointed. That reason could be a poor experience with a product, a support function, a company policy or even a PR statement like a blog, article or advertisement.

Getting to the heart of the issue is paramount for the success of these CX/VoC teams. You don’t have to be J. Edwards Deming to know that identifying the actual problem is the most important step in resolving it. With that said, why is getting to root cause so difficult?

Root Cause… just saying it out loud feels so 1990s. I could offer up a whole dissertation about TQM or “The 5 Whys” but I won’t because everyone reading this blog knows that root cause analysis requires deep drill down well past the surface. But that’s really the issue facing CX/VoC teams today – how does one get past the high-level themes, that are paraded around today as insights, to find the true root cause of customer issues? When I say the ‘true root cause,’ I mean something so concrete that action and accountability can be tagged to it. A team can only affect change if it can identify the root cause of an issue and the proper course of action to resolve it. Otherwise, they’re left discussing “interesting” findings, or focusing on symptoms of a larger problem; neither of which will result in positive change.


Does your company rely solely on survey data to gauge the customer experience? Download our whitepaper to learn the difference between customer experience and customer feedback. 


What does root cause look like through a CX/VoC lens? Typically, CX/VoC teams are looking for macro-level trends in customer interaction data to detect inflection points that can be indicative of a change or problem. The next step is to review the metadata over that timeframe to see if there is a particular call or chat type, social channel, customer cohort or marketing campaign that is driving the change. Next, the unstructured data from the interactions must be evaluated for contextual information. This can consist of sentiment analysis and specific topics that are tagged to interactions. This aggregated data can reveal the “what” and the “why” of the trending issue. The “what” is usually either a specific product/service or a support function such as billing, shipping or digital. The “why” is the actual problem the customers are encountering. This could be a defect in a product, a service disruption, an inaccurate bill, missing shipment, a problem on the website, etc.

Analytic tools that will provide the level of granularity needed to identify the true “what” and “why” are the key to getting to this point. It’s one thing to know that customers are complaining about shoes falling apart, but the path to true root cause requires knowing the model of the shoe, the part of the shoe that is defective (sole, toe box, heel, etc.), the specific material that has the problem, the description of the problem (rip, tear, falling off, etc.) and even the usage of the customers (serious runner, casual wearer, etc.). That’s the context that is necessary to begin framing root cause.

After understanding the model of shoe, part of the shoe, material, description, customer usage, and the number of complaints matching the criteria at hand, the CX/VoC team is ready to hand off the data to a product team that can identify the actual flaw in the shoe’s design and make the necessary changes.

As stated above, it’s important to have the right analytics software to make this process fast and repeatable. Topbox’s Conversation Analytics platform aggregates conversations from every channel and classifies them with rule-based and AI techniques to provide that level of granularity that’s so critical to root cause analytics. If your CX/VoC team is struggling with root cause, give us a call!


Mid-Market Companies CAN Run with the Big Dogs When it Comes to CX Strategy and Execution

Mid-market companies face the same pressure to meet customer needs as their multi-national enterprise counterparts. In fact, one could argue that the pressure is even greater on mid-market companies given the fierce competition that they face. Commonly defined as companies with revenue between $100 million and $1 billion, many of these companies differentiate on service or a niche they fulfill. They often aren’t the lowest cost or fastest providers, which means customer loyalty is paramount for these organizations. While many have worked tirelessly to establish themselves in the market, they’re finding the competitive pressure more and more difficult to overcome. So how can these mid-market companies continue to differentiate themselves enough to earn loyalty and new customers? By gaining a deeper understanding of their customers through the conversational data that’s being stored on servers every day.

Big data and associated analytics can seem daunting for companies of any size. Even the largest companies struggle mightily with complex data integrations.  Adding NLP, natural language processing, to the mix (all the conversations happening across disparate communications platforms) makes it even more difficult. Here’s the good news: a new generation of “born in the cloud” technology, like Topbox, makes it not only feasible but practical for companies big and small to leverage all untapped conversational data in call recordings, chats, emails, survey responses, product reviews, social media and more… Imagine being able to join and classify all conversational data consistently and accurately for deep analysis. In some cases like Topbox’s, these SaaS solutions can be PCI and SOC type 2 certified, and even HPPA compliant taking security concerns off the table.


Create a CX program that’s scalable, measurable and drives revenue. Download this ebook to learn how to turn CX insights into action.


Once all the data has been aggregated, normalized and classified in a secure environment, now what? Mid-market companies don’t typically have an army of data scientists or analysts standing by to comb through millions of customer interactions.  Again, there’s good news. Clunky, bolted-together interfaces that have precluded the common person from interacting with insight-rich data sets have been replaced with elegant user-friendly designs which enable users of every skill level to leverage insights.  From the CMO to contact center supervisors, it’s not uncommon to see a multitude of mid-market clients using Topbox. Getting the insights into the hands of business units that can actually affect positive change is a sure-fire way to continuously enhance the customer experience, earn the loyalty of customers, and create awareness in the market.  A CX strategy that focuses on identifying and fixing customer friction points will win at any scale.

Cost is the next hurdle to jump.  Even with clear ROI, getting op-ex or cap-ex approved for software can be difficult.  Multi-year licensing can demand upfront cap-ex in the millions and leave companies stuck with a version of the software for the term of the contract.  Here’s where SaaS really gets interesting for mid-market companies.  Subscription pricing based on consumption (in this case the number of interactions processed) makes solutions like Topbox affordable for the mid-market. Companies can choose which channels to include and how many interactions to process on a monthly basis in order to control their costs and ensure they get the customer experience insights they need to make improvements. Further, with a SaaS model, these companies will access all the latest software enhancements that providers are releasing.

There you have it, mid-market companies.  The hottest CX tech – deployable, scalable, usable and affordable – is available for you to build upon your competitive advantages that you’ve worked so hard to create!


How To Report CX Insights To Ensure Company-Wide Adoption

“How do you report CX insights to ensure company-wide adoption?” This is a great question and there is no single answer that fits for every company; however, there are some guidelines and strategies everyone can use as they transform their company into one that is focused on CX.

Delivering insights is a crucial step for a CX program. For most programs, this will be the first time the broader company gets a chance to see all the amazing work you have accomplished. Every company has a different preference on how to deliver data and insights. Some companies require high-quality agency level work, while others only need to excel spreadsheets. Regardless of the format, delivering insights, organizing your work and telling a cohesive story is crucial.

Your customers go through a multitude of experiences, and they have talked about them all. This means you probably have insights on every single one of their experiences. In order to ensure company-wide adoption of the CX program’s insights, it is best to group the insights by a subject, theme, or owner. These groupings can be based on business function, experience, products, services, season, membership, or any customer demographic data you have. The goal is to make the report concise and easy to understand. Executives are more likely to read your reports if it can be digested quickly.

When building CX reports, try to limit the number of insights to 6-8 per theme. When too many insights are included in a single report, recipients tend to get insight fatigue. Focus on the most impactful insights that will garner the most attention.

Need help creating and scaling a CX program? Download our ebook to learn how to build a CX program from the ground up. 

When reporting insights, you may run into a team that is reluctant to accept your findings. In these cases, it is best to provide these teams with insights that are familiar to them – even going so far as to give them insights they already know. This will get them to buy in on what you are uncovering. Once you have established a rapport, then you can slowly start providing them with insights that don’t align with their business plan.

Along with insights, CX reports will need to have action items that coincide with specific KPI’s. Most analytics functions are only responsible for gathering insights. In regard to a CX program, however, you are tasked with not only gathering insights but identifying who can take action on them in order to improve the customer experience. A CX program should not only be insights-driven, but results driven.

The CX reports should include the following

  1. Title – What is the focus of this report? Provide a high-level theme or subject.
  2. List – List the insights, current status, teams impacted, measurement KPI, and priority level. Remember, less is more when it comes to impactful insights!
  3. Customer Verbatims – Give 2-3 customer verbatims for each insight. Customer verbatims make the experience more real to recipients.
  4. Details – Go into detail of each insight showing more customer verbatims, the insight(s) identified, and the actions that will be taken. This is a good place to display any graphs or charts.
  5. Wins – If there are any customer experience improvements that have been actioned on from previous reports be sure to celebrate them! This is where you give the proper business functions credit for their work.

The last objective of a CX program, in relating to reporting, is to distribute the report to as many people as possible. It’s in the best interest of every employee to know and understand the voice of the customer. You never know who is going to read the report and find the insights meaningful!

Analyzing Non-English Interactions

It has been estimated that less than 5% of the world’s population speaks English as a first language. Even in the United States, over 20% of the residents speak a language other than English when at home. At the same time, businesses are realizing that, in a world where consumers have more choices than ever, the customer experience is the single greatest differentiator. According to Forrester, 72% of businesses name improving customer experience as their top priority. But are non-English interactions with your customers left out in the cold?

OK, maybe you’ve jumped the first hurdle and hired multi-lingual agents to handle those interactions. You might even have a small group of multi-lingual QA specialists to review a handful of those calls and chats, but their efforts will be focused on contact center performance. According to Gartner analyst Augie Ray, “a company’s CX focus can be narrowed to activities like enhancing customer care processes, front-line employee performance, customer care hiring, and call center training. While these are all good and necessary efforts, this sort of myopia misses the point of what CX really is and what it can do for your company.”

Whether the focus is serving non-English speakers in the United States or across multiple geographies, understanding the issues and needs of these segments is an absolute necessity for a customer experience program. One size doesn’t fit all! Marketing departments, product management teams and contact centers can all benefit from insights derived from non-English customer interactions. So how do you know what’s really being discussed on all those Non-English interactions? More importantly, how can you mine those interactions for actionable insights?

Topbox has woven accurate, automated translation technology into the fabric of a highly-sophisticated Customer Experience Analytics platform. Multi-lingual calls, chats, emails, surveys, reviews, social media, and more can all be translated into English and analyzed for CX insights alongside the rest of your customer interaction data. It’s part of Topbox’s proprietary data normalization process that aggregates structured and unstructured data from every channel, and applies a classification model unique to your business.

Whether the interaction happened in the call center or on social media, in English or Mandarin, customers from all walks of life are giving you actionable feedback. It’s more important than ever that you listen to understand their experiences with your company. Now you can… with Topbox.

The Difference Between CX and CFM

The Difference Between Customer Experience and Customer Feedback.


Most companies have deployed one or more technologies to capture customer feedback: NPS Surveys, CSAT Surveys, Point of Sale Surveys, Product Reviews and more have been around for decades.

However, companies that rely on solicited feedback like this as their only source of customer experience data are going to be left behind. Your customers already give you rich, actionable feedback thousands of times a day, and you don’t even have to ask for it.

Download this paper and learn how Customer Experience Analytics is helping businesses understand the voice of the customer and improve customer experience.

Get the Whitepaper


Survey Says…Time to Move on From Survey Scores

Companies have been using surveys to try and understand their customers for decades. Similar to the game show Family Feud, surveys are a fun, lighthearted way of understanding people that shouldn’t be taken too seriously; despite this, companies continue to put serious weight behind their survey scores. Business decisions are continually made based on a handful of customer surveys. Executives monitor NPS and CSAT scores that rarely move – and even when they do fluctuate, the scores provide no context for what caused the change to occur in the first place. Surveys can have value but relying on them as the core KPI to represent your customers is extremely dangerous. Customers no longer want to be asked questions, they want to be understood through their interactions with your company.

The most popular survey scores used today are NPS and CSAT. Both scores were created to gauge the loyalty and satisfaction of a company’s customers. While these metrics are used in a variety of ways today, they generally aim to append a score to a customer’s experience. In the new relationship economy where customers care as much about the experiences as the products they are buying, having a quantitative measurement associated with customers is crucial. The underlying need for this measurement will always be a top priority, so having these surveys in place makes sense.

The problem with these surveys is that customers are likely giving a score based on a specific element of their experience, say the end of a conversation. What is missed is an analysis of the whole customer experience – from start to finish – which often tells you why they contacted your company in the first place.


Does your company rely solely on survey data to gauge the customer experience? Download our whitepaper to learn the difference between customer experience and customer feedback. 


Customers contact your company because they experience an issue with your brand. This could be related to a product, service, or digital experience. Regardless, the customer experience didn’t go as expected. Most companies resolve these broken experiences with speed and professionalism, satisfying the customer who in turn gives a high score on a survey. When the high survey scores are reported to business functions and executives, everyone celebrates and congratulates one another. Why wouldn’t they? Customers are seemingly very content with their experience.

In reality, there continues to be customer experience issues upstream that are forcing your customers to contact you and take the survey. Although your company may be receiving high survey scores because you’ve done an excellent job solving customer issues, those survey scores may be masking more problematic customer experiences that can drive customer churn.

Topbox has analyzed customer interactions for some of the largest brands in the world and found that almost half of all perfect survey scores originate from a conversation that had to do with a broken customer experience. In addition to identifying where the customer experience went wrong, Topbox can pinpoint what ultimately drove the customer to become satisfied and submit a high survey score.

Surveys are a legacy way of understanding your customers and should only be used as loose directional data. They rarely provide enough context and aren’t always a good representation of your customer base. Rather than ask questions, you should learn to infer what customers are feeling during every customer interaction. If you have an issue with your friend or spouse do you ask them to take a survey so you can better understand them? No, you communicate and learn about one another through dialogue. Companies should be treating their customers in the same way. With a solution like Topbox, you can not only analyze the sentiment of an entire conversation but infer what topics and phrases are driving that sentiment. Conversations go through a rollercoaster of highs and lows and can’t be tied to a single score.

Companies should be analyzing interactions across every touchpoint that customers engage businesses: chat, call, social, reviews, blogs, etc. Surveys can’t be used at every customer interaction, so inferring how customers feel is extremely difficult without the right CX solution. It is not logical to expect your customers to give you feedback every time they interact with your company. For some customers, simply asking them to take a survey diminishes the experience.

Surveys were created during a time when analyzing customer conversations and actions was not possible. With capabilities like Artificial Intelligence (AI), Machine Learning (ML), and Natural Language Processing (NLP) that are accessible today, solutions like Topbox enable companies to move beyond their survey data.

While surveys can be great when used at a high level, they should never be used to make major business decisions. When your company communicates with its customers, be sure to understand the entire conversation, not just a portion of it. You’ll be amazed at how much you can learn when you listen.