Analyzing the Analytics


Analyzing the Analytics

Analytics, namely the analysis of data such as through software tools to obtain meaningful and actionable insights, is central to the contact center. And contact centers collect a bounty of information from a growing array of channels and sources.

Understanding customers and contact center agents, and their interactions and experiences from this data via analytics, enables organizations to provide higher quality customer and agent experiences while driving in more revenue and lowering costs.

Given the vital importance of analytics, Contact Center Pipeline has reached out to leading suppliers to obtain their insights into these solutions, including their uses, trends, and future, along with recommendations to maximize the value of these investments.

We have had virtual conversations with Sharon Einstein, Vice President and General Manager, Customer Engagement Analytics, NICE; Jeff Gallino, CTO, CallMiner; Terri Kocon, Product Marketing Manager, Calabrio; Tibor Vass, Senior Director of Industry Design – Retail Solutions, Genesys; and D. Daniel Ziv, Vice President of Speech and Text Analytics, Global Product Strategy, Verint.

Q. What changes have you seen in the requirements from contact centers for analytics solutions over the past 12 months?

Sharon Einstein

Sharon Einstein: We’ve seen a spike in demand for customer experience analytics solutions that focus on digital interactions.

Specifically, there’s been high demand for solutions that can take insights from voice conversations and apply them to digital self-service interactions.

Businesses are determining the reasons that customers are having to speak with live contact center agents and then making sure they have appropriate coverage in their self-service channels to resolve those issues.

To do this, they are analyzing the interactions from high-performing agents, identifying the tasks, activities, and even the languages those agents use to resolve these intents and using that training data to program self-service applications. Ultimately, the goal is to take the intelligence from conversational data generated by agents to rapidly build smart self-service that delivers improved experiences across digital channels.

We are also seeing a twofold change when it comes to the agent experience. First, contact centers are looking to capture the behaviors that drive positive experiences, such as “was the agent empathetic?” or “did they build rapport or take ownership of the conversation?” and not just the checklist activities.

Contact centers are looking for elements that take place continuously throughout the calls and are tapping into analytics and artificial intelligence (AI) to obtain that data easily and at scale.

Capturing those behaviors and evolving quality and performance management programs to be more behavior-focused is certainly a top priority for contact centers.

Second, we’re seeing the need to prepare agents to handle higher-stakes conversations as consumers tend to be more emotionally charged after waiting in the queues for long periods of time or not getting the answers they’re looking for online.

Also, brands are investing more in their front-line operators to make sure their brand customer experience (CX) is being used as a competitive differentiator as well as to boost employee engagement and combat the Great Resignation.

They’re providing real-time interaction guidance on what is the next best actions, behavioral cues, or recommendations to agents, so they have real-time assistants that are helping them better manage these interactions.

Jeff Gallino

Jeff Gallino: Let’s look at these developments from (a) the CX and (b) the employee experience perspectives.

With (a) we’ve seen a lot of changes in CX over the past two years. The first year of the COVID-19 pandemic was marked by organizations quickly working to adopt the technology solutions needed to successfully transition their contact centers to fully remote operations.

But the last 12 months have been more about using those solutions to better engage with the customers on their preferred communication channels.

More customer interactions are happening across more channels than ever before. These omnichannel conversations hold a wealth of insights. And more organizations are realizing that if they capture and analyze these interactions (voice, text, emails, chats, etc.), they’ll be able to uncover key insights that make it possible to take action and improve CX.

The best CX analytics solutions are those that can capture and analyze 100% of customer conversations, regardless of channel. They also have to be agile, scalable, and easily integrate with existing technology solutions, such as CRM systems.

These attributes make it possible for organizations to gain holistic views of their customers’ journeys, better understand the pain points, gaps, and trends across all customers, and identify areas of improvement for CX and customer outcomes.

With (b) the past 12 months have also proven that the shift to remote or hybrid work is permanent, and contact centers are going to have to continue supporting a distributed workforce of agents.

The analytics tools that are going to help contact centers succeed in this remote future will be those that improve agent performance and lower their turnover, increase operational efficiency, and more effectively manage costs, among the goals.

In the contact center of yesterday, agents and supervisors worked in the same rooms, where supervisors could listen in on interactions, give agents tips in between calls, and conduct in-person team building meetings.

Technology, such as analytics solutions, has now had to replace all of the onboarding, training, coaching, and ongoing performance improvement efforts that were previously done in person.

The best contact center analytics solutions offer a range of features, including real-time feedback, such as next-best-action guidance, post-interaction insights for individual and team performance, and two-way collaboration capabilities.

These solutions boost agent productivity, engagement, and motivation. When agents feel supported in their roles, they deliver better service and experiences to customers.

For supervisors, contact center analytics solutions make it easier to identify performance trends, offer behavior reinforcement or coaching as needed, share best practices from top agents across the teams, and maintain a culture of persistent improvement.

These features are no longer nice-to-haves; they are the only ways to successfully operate contact centers in today’s remote world, while meeting the needs of agents and supervisors.

Terri Kocon

Terri Kocon: In response to the COVID-19 pandemic, the Great Resignation, and the growth of the cloud, analytics has rapidly evolved over the past 12 months.

There have been two main developments for contact center analytics solutions. First, analytics is increasingly being used in broader ways than just traditional contact center quality assurance. It is also now being used as a primary tool in the pursuit of the connected enterprise.

In Calabrio’s 2020 State of the Contact Center Report, we found that contact center managers are recognizing an intense and growing demand for analytics to help businesses take action. The report also revealed that 90% expected analytics demands from the wider enterprise to grow, especially from the executive suite.

The second main development in analytics is the availability of cloud solutions. Calabrio research has shown that cloud-based solutions are directly helping contact centers meet the challenges of the new era—and that includes facilitating analytics needs.

In fact, nearly 70% of cloud-powered contact centers report that cloud-based solutions are helping them improve their analytics capabilities for both customer and employee data.

Cloud-powered contact centers can see more value in their employee and customer data analytics as it is more accessible and achievable. The cloud helps facilitate analytics across all channels, including voice, text, email, and social in this new omnichannel world, giving contact centers vital visibility into them.

Tibor Vass

Tibor Vass: Modern customers are choosing between brands quite differently than they did in the past.

Back then, product quality, price, and reliability drove buyers’ preferences. But now the social and human aspects, like word of mouth, peer reviews, service excellence, empathy, ecological footprint, and inclusion are equally—if not more—important than the products’ capabilities.

Also in the past, analytics were primarily used to serve the sellers’ interests and were designed to improve sales directly.

But with the shifts in consumer behavior, brands are using analytics to match their offers with buyers’ preferences and real-time intent, developing win-win selling scenarios that builds trust and loyalty. This is especially true in the subscription economy, where the one-time sale is just a fragment of the customer lifetime value.

CX is therefore not a one-time thing—it’s a collection of experiences over time across many ecosystem and journey components–and it needs to be orchestrated in a predictable way for each individual customer.

Brand representatives—regardless of their marketing, sales, or service roles and whether they are robotic or human (like contact center agents)—need systems that can leverage analytics in a fully automated manner.

Analytics should be leveraged to support the customer conversations through functions like agent assist, conversational scripting, and next-best answer/next-best offer recommendations in real time.

Historic and real-time analytics plays key roles in quality assurance, performance improvement, and continuous learning. They need to be integrated in all journey steps across the modern CX ecosystems to perform continuous improvement autonomously.

D. Daniel Ziv

D. Daniel Ziv: With the acceleration of digital transformation, simpler tasks are being self-served via digital channels, leaving the more complex and emotionally charged interactions to be handled by live agents in the contact center.

These developments are driving up handle time and cost and are presenting challenges for agents who are required to provide more empathy to meet customer expectations while handling challenging interactions. This has driven demand for solutions that can measure customer sentiment in real time and identify opportunities to improve CX during the interactions and not just after the fact.

At the same time, working from home or anywhere has shifted from a temporary solution to a permanent strategy, in part because hiring capable contact center agents has become more challenging, and organizations are required to offer the flexibility of remote work to attract the best talent.

However, working remotely creates challenges for managers and supervisors who have less visibility into agents’ performance, and fewer options to provide guidance, support, and coaching.

These needs have created demand for AI solutions that can monitor and improve agent experience and performance in real time. They can provide the necessary visibility and real-time alerts as well as guidance for agents especially during moment-of-truth interactions such as complaints, escalations, negative sentiment, extended silence, interruptions, and crosstalk.

These solutions can also drive contextual knowledge, providing the information and answers to increasingly complex sets of policies, regulations, and enabling compliance with them.

In addition, they provide positive reinforcement when agents appropriately provide the right levels of empathy and thus drive positive customer sentiment.

Q. There has been an evolution in customer contact channels like the advent of asynchronous/social messaging platforms and the coming of age of video. How have these new channels impacted, or will impact the ability of contact centers to obtain meaningful, critical insights from customers and agents?

Sharon Einstein: Looking at the full customer journey, interactions start with the consumers’ most likely first activities, like Google searches, which would extend to voice and digital interactions. Brands want to know how to make those journeys as seamless and as informed as possible to enable consumers to self-help.

Being able to use analytics to see where there are bottlenecks, challenges, and distress points throughout those journeys that are not making those experiences flow is one of the main priorities of brands today.

There is also a huge opportunity to better utilize the technology within the devices we use and interact with every day (e.g., our smartphones) to identify and resolve problems faster and more efficiently.

For example, consumers activating their cameras to allow agents live looks into what they are seeing, such as perhaps a problem with their streaming service, and capturing visual data such as QR codes. Bottom line: the more data you have on this front, the better insights you’re going to be able to derive from those data sets.

Jeff Gallino: While there has undoubtedly been an expansion of customer contact channels over the past few years, most contact center analytics programs remain rooted in the traditional voice (phone) and text (chat, email, etc.) data. In fact, research indicates that contact centers saw unprecedented spikes in call volume during the pandemic.

Videoconferencing is a great example of a new channel that’s still fully dependent on voice as a data source.

While we’ve seen a pandemic-driven boom in platforms, such as Zoom and Microsoft Teams, voice is the key part of these interactions; meetings can continue without video, but not without voice.

That said, new interaction channels like social media and video present opportunities to access deeper customer insights and unique behaviors.

Instead of hindering a contact center’s ability to uncover meaningful insights, these channels provide added layers of data and paint more complete pictures of customers’ interactions at scale.

For example, threads of social conversations or messaging platforms (like Facebook Messenger) can be ingested into analytics platforms. They can inform customer trend analysis, customer journey mapping, and customer emotion/sentiment alongside traditional voice calls or text functions.

Further, advancements in customer journey analytics—and deeply understanding conversations across contacts—will, in turn, lead to a revolution in asynchronous messaging support.

There still needs to be progress made in the capabilities of video analytics. But as those technologies advance and can glean nuanced information from faces and reactions, analytics solutions will incorporate this additional layer of meaning to understand individual customers’ emotions during contact center conversations.

Terri Kocon: There’s no doubt about it, the omnichannel contact center is here. Modern enterprises are inundated with data and the pace of information flow is only accelerating.

To complicate matters more, all the data that pours into an enterprise isn’t valuable, but much of it is essential. Parsing out the important data from the white noise has then become a vital enterprise need.

And that’s where the contact center comes in.

The contact center is the one department that touches every customer communications channel and can act as a centralized source for all customer data.

By leveraging tools, such as business intelligence (BI) solutions built for the contact center or those that enable the importation and analysis of web and video meeting recordings, contact centers can analyze all interactions across every channel and use that information.

These solutions can be used not only for traditional quality assurance, but for actionable insights related to other topics as well, such as product marketing and back-office questions.

The biggest challenge facing the contact center today is transforming the raw data into meaningful metrics. The enterprise that can do that will succeed in the new era.

Tibor Vass: Over the past two years, digital channels, asynchronous messaging, and video capabilities have evolved as important contact center channels.

Social channels have always had a key role in marketing, but now social is playing a significant role in sales, and is becoming more important in customer service interactions as well.

Social listening was automated in the early phase of digitalization of businesses, but how to translate social analytics to actionable insights remained a challenge.

Asynchronous messaging—just like social conversations—are particularly important and convenient for customers because these interactions happen in non-real-time where the customer is driving the timeline and the channel preference according to their own convenience.

For brands to follow each individual customer conversations asynchronously also presents multiple challenges. The main one exists within technology systems that need to follow the customer context across “random” channel switching in an asynchronous manner.

At the same time, brand representatives need to be comfortable with, and capable of, picking up individual conversations wherever these consumers left off.

This critical task requires intelligent systems, real-time integrations, and well-trained human and context-aware robotic resources to cope with that challenge.

Conversational AI, analytics, process automation, and agent assist systems need to be working in perfect sync to analyze data, capture context, predict intents, and deliver real-time recommendations in a fully personalized manner without delay.

D. Daniel Ziv: Asynchronous messaging channels are far more automation-ready than nearly any customer contact channel that has come before them. Apple, Google, Facebook, Tencent, Twitter, and others have made huge investments in building out the bot capabilities of their respective messaging platforms.

This has made messaging an extremely attractive place for brands to build conversational automation that addresses the most common customer inquiries: while reducing inbound volume for live agents in the contact center.

That said, the combination of bot and agent-based interactions as well as the blend of real-time and asynchronous interactions presents some challenges when it comes to obtaining meaningful contact center insights.

While conversational AI has proven to yield compelling efficiency gains, impressive reductions in cost per contact, and clear improvements in customer satisfaction, it has also created a challenge for brands to get holistic views of their customer engagement performance.

Contact center leaders need to be able to fuse bot-based conversation data with data from conversations handled by human agents to gain accurate insights into customer experience quality.

To accomplish this, brands need solutions that capture the “effort” across bot and agent engagement in a manner that allows brands to make apples-to-apples comparisons between asynchronous interactions with those that are linear, like calls, chat sessions, or video concierges.

These solutions also need to factor the nuances of asynchronous conversations into their analytics, tracking interactions that may start with bots, get handed off to humans, and back again, until the customers’ intents are addressed.

Q. What analytics tools should contact centers consider employing to meet their needs going forward? And what best practices can you recommend in selecting, deploying, and using them to maximize the value of these investments?

Sharon Einstein: With AI coming of age in its use and application, I recommend brands move away from tools-based strategies where appropriate and start incorporating more AI solutions.

AI and machine learning are performing the human activities of using analytics tools, but they’re doing it in a way and to a quality level and scale that humans can’t achieve.

I recommend opting for a “complete” AI solution and avoiding siloed applications.

One that handles everything – from that original Google search to interactions across digital and voice channels to the agents – providing the visibility and exposure to derive insights across those channels. Rather than something that’s limited or dedicated to one space.

Jeff Gallino: When choosing a contact center analytics solution, organizations need to balance finding the right platforms with finding a vendor that is going to operate as a true partner.

On the product side, the solution should have the features and capabilities that are most important to their businesses. This includes the ability to capture and analyze omnichannel customer interactions at scale, and improve operations and agent performance.

But perhaps most importantly, the solution should uncover insights from customer conversations that drive improvement inside the contact center and across the enterprise.

Being able to not only capture massive amounts of data, but help organizations make sense of it through actionable insights, is a huge competitive differentiator.

To find a true analytics partner, choose a vendor that knows the space inside and out (and maybe even pioneered aspects of the technology), and which has one foot in the future.

The vendor should be committed to continuous innovation and product updates, while also being financially stable enough to keep organizations’ unique business goals top-of-mind regardless of any consolidation activity—which is inevitable as the analytics market continues to boom.

This partner should work to understand specific business goals and objectives to accelerate ROI and business improvement.

Instead of getting through implementation and going quiet, a vendor that takes a comprehensive and consultative approach from deployment through the lifetime of the partnership ensures alignment for ongoing success.

Terri Kocon: Contact centers need to consider modern speech analytics with predictive model tools that are designed to meet the needs of the digital-first consumer age, in addition to BI solutions.

In the modern consumer mindset, engaging with companies is less appealing than ever, and so contact centers can no longer just sit back and wait for calls. Assessing the customer journey to preemptively determine where common questions arise enables the modern, proactive strategy that will lead to better CX.

This process can also help contact centers protect themselves from overflowing calls, providing customers with self-service when they want it by mapping customer journeys and predicting self-service needs. With speech analytics, enterprises can build predictive models to illuminate the customer journey and help map out where service may arise.

Another area growing in popularity is sentiment analysis tools. These can help enterprises take advantage of customer interactions by analyzing how customers actually feel about their brands, service, and competitors. The insights are not limited to the words customers choose, but the tools can use AI to reveal whether their thoughts are positive or negative.

Finally, when choosing a new analytics tool, the best practice is to start small and simple. Don’t try to boil the ocean. Look for a straightforward, user-friendly and smart analytics tools and use them for informed results.

Tibor Vass: Brands need to use robust analytics, but the analytics tools need to be aligned with a clear business purpose, and they also must be integrated and orchestrated appropriately.

Individual analytics solutions may be able to improve individual experience moments, but without connecting them into an experience flow they might not deliver the desired result.

Imagine if brands deploy a clever conversational AI bot on the voice inbound side that can understand customer intent and is even able to resolve certain types of queries autonomously based on knowledge base automation.

But, when the customers switch channels (e.g., sending text messages to the brand regarding the same topics), companies often drop the ball, and these buyers do not receive responses for a day or a week.

Customers are omnichannel by nature through their smart devices. Each of them has their own experience flows that crosses multiple touchpoints, leverages multiple communication channels, but which involves multiple brand representatives across their journeys.

Customers’ interactions are becoming increasingly asynchronous and contain multiple intents and transactions at the same time. They require an autonomous analytics and engagement orchestration engine in the middle to provide a frictionless and consistent CX.

But having the relevant analytics engines and large amounts of data are not enough. These systems need to be connected across the CX moments and must be integrated with a customer and an employee engagement engine that can help the brand execute the actionable insights in real-time.

D. Daniel Ziv: Organizations that don’t currently have access to near real-time speech and text analytics should deploy them first, so they can have full visibility of 100% of their customer interactions.

Those that have deployed these analytics solutions should consider deploying real-time agent assist solutions that provide real-time guidance to help support agents and drive optimal interaction outcomes.

Providing real-time guidance to agents effectively requires full context of the live interactions, analyzing what’s being said, the conversational flows, as well as additional context from the agents’ desktops, such as customer profiles and the agents’ actions.

Organizations that are considering deploying real-time agent assist solutions should select those that provide that full context and whose vendor has proven success with other customer deployments.

Organizations should potentially start with a pilot with several hundred agents who volunteer or agree to receive real-time assistance, incentivizing the agents as their performance improves.

Organizations should also avoid too many real-time triggers or confusing gauges in order to allow agents to focus on listening and communicating with the customers. Instead they should select five or six proven real-time guidance triggers to start with. After the initial success, they can expand to additional custom triggers, making sure to select a vendor that can support them as well.