Managing the Data


Managing the Data

If customer interactions are the lifeblood of organizations, pumped through their customer engagement hearts, whose muscles are live agents and automated applications, then data are the individual cells, carrying the vital information the organization needs to thrive.

Heather Richards

Effectively managing that data, so that it gets to where it has to go, and without blockages or other issues, like security that has potentially serious consequences, is key to corporate wellbeing.

To get a pulse on what is happening with data management through the contact center, we recently virtually interviewed Heather Richards, who is Vice President, Go-to-Market Strategy, Verint.

Q. What are the top data management challenges being faced by contact centers, and what are their drivers?

One of the biggest data management challenges being faced by contact centers is data unification and access.

As the customer journey becomes increasingly complex, organizations are grappling with unifying, managing, and accessing huge amounts of siloed customer engagement data. This is further complicated by the fact that it’s more than just one channel and one department accessing and monitoring the data. All of these factors make it difficult to get a full 360-degree view of customer and workforce data.

Business leaders need to access and harness customer engagement data directly in a consumable way, without the need for reliance on outside sources or IT to provide the insights needed to make data-driven decisions.

In response to these challenges, businesses need to look for data hub solutions that ingest, organize, and manage all their engagement data regardless of source. These applications must have artificial intelligence (AI) and language capabilities to provide business users with no-code, self-service access to key insights from their engagement data.

“Greater visibility into
customer interactions across channels is undoubtedly invaluable to brands. But at the same time it can present additional challenges…”
—Heather Richards

Gone are the days of struggling to figure out how to unify, manage, and access large amounts of siloed customer engagement data. Today, the cloud, Generative AI, and the no-code/low code evolution is dramatically changing all of that with smarter user interfaces for data retrieval and visualization, while propelling positive user experiences to the next level.

Q. There are new and increasingly popular synchronous and asynchronous digital channels that customers are using. From a customer and also an employee engagement perspective, do these new channels present benefits but also challenges in extracting, managing, and integrating meaningful data from them?

Customers are using an abundance of new digital channels to interact and engage with brands that are not just limited to video and messaging apps, but chatbots and more. The question becomes: “How can businesses extract, manage, and integrate meaningful insights from customer interactions across all channels beyond voice?”

Greater visibility into customer interactions across channels is undoubtedly invaluable to brands. But at the same time, it can present additional challenges because the data from a customer watching a video or conversing with a bot will arrive in a different form and with different metadata than a voice recording.

Capturing customer data is paramount, but if you aren’t doing anything to transform it into a normalized, usable state, it makes it very difficult to do anything other than store it, and that’s not the point of capturing engagement data. You want to turn insights into action, not just collect data.

Q. What effect has the advent of AI, including ChatGPT in its evolvingiterations, had and will have on contact center data volume, value, accuracy, actionability, security, and management?

The more customer engagement data you have, the more sophisticated (and better) your AI models will become.

AI needs access to large amounts of data, and additional AI automated interactions increase data volume. You can’t have one without the other. Does this influence contact center data volume? Yes, but that’s ultimately a good thing as long as you can harness and utilize that data sensibly.

One of the reasons we’ve seen evolving iterations of AI – such as ChatGPT – so quickly is the availability of vast amounts of data coupled with processing power. Advancements in large language models (LLMs) or Generative AI are made manifest thanks to masses of data.

When it comes to actionability, this again benefits from having a way to access and visualize actionable information from your engagement data. Data hub solutions, coupled with engagement insights applications, allows users to retrieve actionable data insights using natural language queries.

AI is delivering unparalleled visibility into customer engagement operations, transforming raw data into meaningful visualizations and metrics, and offering guidance to assess and improve agent performance as well as customer experience (CX).

There has been a significant paradigm shift, where you see AI being embedded into customer and employee workflows.

Leading-edge suppliers have multiple AI powered bots within their customer engagement applications that do everything from interaction containment, to forecasting and scheduling to knowledge retrieval.

“The purpose of AI is not to
replace humans; instead, it’s augmenting humans by giving them superhuman capabilities via AI embedded within a contact center’s existing applications and workflows.”

These recent developments enable real-time intelligence, predictive AI, and ultimately customer experience automation. And you don’t have to be a data specialist or an AI expert to capitalize on what’s happening, which is exciting.

Regarding security, it is important to understand that there can be a distinction between the data security available from many of the Generative AI tools used in the public realm versus its use for business applications.

LLMs that are deployed for business applications will sometimes use a combination of public data sets and an organization’s own more specialized data in the training of the language models. Safeguards should be in place to ensure that any corporate data is protected from a data security perspective but also remains private, and not used to subsequently train the publicly accessible models.

Of course, new questions must be asked in terms of data access, residency, security, and the use of data for continuing to develop and train the AI models. Vendors and buyers need to look at their own security practices to ensure that the required security and regulatory needs of their customers are met.

Data Management in the Cloud

It has long been argued that data should reside in the cloud for access, capacity, and flexibility, including by remote staff.

So, we asked Heather Richards about where are companies in making that ascent? Are they keeping data on-premise, and if so, what types and moving other data, and if so, what types to the cloud?

And if data storage and access are split, does this pose an issue when seeking meaningful intelligence and insights from it?

Here is Heather’s reply:

“When organizations are utilizing on-premise technologies, this often means data is also on-premise.

“In some cases, there is still the fear of putting sensitive data into the cloud. This fear has diminished substantially over the last few years with more robust security available for cloud applications. However this doesn’t solve the problem of having the requirement for a mixture of on-premise applications and cloud data storage.

“Organizations are not going to migrate all customer applications to the cloud overnight. Many companies may still want to keep things such as ACDs on-premise because it’s too expensive or disruptive to move everything to the cloud simultaneously.

“At the same time, companies are adopting new digital applications and channels such as intelligent virtual assistants that are born and raised in the cloud.

“Organizations are looking for contact center-as-a-service (CCaaS) platforms that allow for a blended approach to keep the infrastructure where it is, but also take advantage of AI infused cloud applications that drive CX automation.

“The blending of applications in the cloud with those on-premise makes an open platform approach so important. And you need the ability to ingest, normalize, and access data regardless of where it resides.”

Q. Contact centers have long had staff turnover, but this factor appears to be becoming more serious with the Great Resignation. Has this issue been having an effect on agents’ abilities to effectively handle data and be proficient in using new data management methods and solutions?

The goal is to embed AI into workflows and processes so that no one needs to have to “figure out” how to use AI. AI becomes the assist that is embedded into existing workflows.

The purpose of AI is not to replace humans; instead, it’s augmenting humans by giving them superhuman capabilities via AI embedded within a contact center’s existing applications and workflows.

The more contact centers can embed AI into existing workflows and processes that its agents are already familiar with, the better. It’s just there, and it will ultimately have a positive impact on agent turnover because now you can assign a bot to tackle and automate repetitive tasks.

For example, some applications have automated call note wrap-up bots, which use Generative AI to provide real-time transcriptions and summaries of them. There are also specialized bots that provide seamless transitions from self-service to assisted service.

Additionally, contact centers can also automate repetitive tasks such as agents having to answer the same questions repeatedly through self-service utilizing Conversational AI and knowledge management.

Bottom line: automation frees up agents to do much more interesting work, which leads to greater job satisfaction and hopefully lower turnover while allowing them to improve the CX by applying their gained knowledge and expertise.

Q. Siloed data has long been a problem, particularly from the customer and employee experience perspectives. So why do too many companies keep it locked up?

Data shouldn’t be (and likely isn’t) being kept locked up intentionally. Rather, the data originated in an application that is not easily or fully accessible therefore creating a silo: and connecting this siloed data can be difficult.

These data silos are why it is important to have the ability to bring data together from disparate channels, interactions, and applications. It is this problem that leading vendors are seeking to solve with a data hub. If this data can reside at the core of a customer engagement platform, it can then be used to both drive actionable insights across applications – but also be used to train the AI embedded into applications and workflows. It then contributes to improved operational efficiency as well as CX automation.

“With an open platform approach, you can harness engagement data from multiple applications or channels: breaking down data barriers…”

Today, it’s more vital to be able to leverage AI and automation at the core to elevate CX across channels. This is ushering in an approach where organizations must reimagine their contact center architecture with an open systems approach to help organizations make engagement data work 24/7, put AI at the fingertips of agents, maximize CX automation, and support best-in-class operations.

In the past, customer engagement applications and data repositories were built in vacuums. With this “closed” approach, data remains locked within a siloed environment.

With an open platform approach, you can harness engagement data from multiple applications or channels: breaking down data barriers where data isn’t locked within a single vendor environment.

You can then normalize and enrich that data so you can draw parallels and connect the dots in a consumable and actionable fashion. Which is again why engagement insights applications are so powerful because they give businesses a way to easily retrieve and find actionable insight in their data: and improve the AI models that are used to drive CX automation.

Q. What types of solutions now exist to help improve data management in the contact center, what are their benefits, challenges, and your recommended best practices for implementation and use?

There are many applications that capture customer engagement data. Call recording, text interactions, self and assisted service interactions, Conversational AI solutions: all of these applications capture data, and to different extents make the data available.

However, as this data grows both in amount and complexity, new solutions are required to allow organizations to find actionable insights within the data and harness it for things, such as AI models, that can be used to successfully automate further customer interactions and drive CX automation.

“You need an approach that enables the normalization and enrichment of all engagement data, to drive better AI models, and provide more CX automation.”

Data hub solutions are designed to help make this task easy and affordable. They enable organizations to bring structure to unstructured data across voice and digital channels and enrich data to drive even deeper insights from it. And those vendors that use an open platform approach eliminate the hard work of customizing them for their buyers.

Organizations need data hubs that have adapters that allow them to freely import and export data, and also have easy, intuitive-to-create data management and compliance workflows to help them bring together, manage, and enrich interactions from a variety of channels and modalities.

Regardless of your application blend, you need an approach that enables the normalization and enrichment of all engagement data, to drive better AI models, and provide more CX automation. You will need to apply the tools necessary to provide actionable insights from this data to business users – not just data analysts – regardless of the complexity of the customer journey.