Google announces general availability of CCAI Insights

Google announces general availability of CCAI Insights 1600 1200 Kane Simms

Google has just made understanding your customers at scale a whole lot more accessible for all businesses with CCAI Insights.

Google announces general availability of Contact Centre AI Insights

Google announces general availability of CCAI Insights

At Google Next ’21, Google announced the general availability release of CCAI Insights. It’s part of it’s wider CCAI offering, which also includes DialogFlow ES and CX, as well as Agent Assist.

What is Google CCAI Insights?

CCAI Insights is aimed at contact centre teams to help them gather understanding on customer interactions. It should help customer service teams make better business and customer experience decisions. It uses natural language processing and natural language understanding to organise and structure customer and agent conversations to extract insights and trends.

When we spoke to the then-Head of Conversational AI and Contact Centre at Google Cloud, Antony Passemard, we heard about the Agent Assist and DialogFlow capabilities of CCAI, but Insights was still being developed. Now, it’s ready for prime time.

CCAI Insights features

CCAI Insights has four main capabilities:

  1. Smart highlighters that highlights important parts of a conversation, such as when a customer makes a payment.
  2. Cloud Natural Language Processing (NLP) Integrations that profile the changing sentiment of agents and customers during a call. And entity extraction that can pull out key pieces of data from a conversation, such as names and dates.
  3. Custom highlighters that work the same as smart highlighters, only allow you to train the system on the specific phrases or intents you’d like to monitor.
  4. Topic modelling which aggregates topics discussed within conversations, so that teams can see why their customers are calling.

Why is CCAI Insights important?

These features solve some of the problems that contact centre teams have had since the beginning of time. They also create entirely new opportunities and possibilities that weren’t even dreamed of before.

Understanding what people are calling about

Most contact centres have manual processes to understand why people are calling. This usually requires agents to physically type or click a ‘wrap’ reason from within the contact centre system. This is always a challenge to complete accurately or even to do at all when you’re under pressure to get to the next caller.

Wrap reasons also tend to record the most salient need a customer has i.e. the one the call centre agent can remember. In reality, customers may have multiple needs during a call.

For example, perhaps they call about making a bill payment, but while they’re there, they ask about their next bill amount, due date and their remaining contract length. An agent would use a wrap code of ‘Bill Payment’, when in reality, there were four needs that were resolved on the one call.

Topic modelling

The Topic modelling feature of CCAI Insights allows you to get at that more detailed data. Then you can recognise patterns and needs that weren’t previously visible to the business.

This is invaluable data for resources planning, automation strategy and also can help inform the wider business of customers needs to shape content, communications and marketing materials.

Quality control and agent consistency

Another age-old problem is creating a consistent service level across business units, departments, tiers, teams and people.

Perhaps one agent is mighty-pleasant while another is having an off-day. One agent might do one thing consistently well, where another might overlook.

Being able to find examples of successful, model calls to use for training, and to find the conversations that didn’t go so well to improve, are both important in making sure that customers receive consistently high standards of service.

Typically, this would require a call centre manager to painfully listen back to a sample of calls. Finding the bad and good ones among all of the thousands or millions of calls you receive per year isn’t exactly easy. Or they’ll shadow agents and listen in on calls, which isn’t great for agent confidence or morale. There’s nothing like working while someone’s looking over your shoulder.

Smart highlighters

With Smart Highlighters, CCAI Insights will automatically highlight important parts of a conversation, such as when someone has been authenticated, put on hold, successfully completes a purchase and others. This will enable managers and teams to quickly sort the wheat from the chaff and find areas of importance within calls.

Data input

Agents have to do a lot of multi-tasking when on calls with customers, switching screens and inputting data into multiple systems, all while listening intently, responding appropriately, maintaining a structured conversation and, in the back of their mind, trying to meet their KPIs.

While CCAI Insights doesn’t specifically solve this problem, it does have capabilities that could do in future.

Entity extraction

The Cloud NLP integration capability is able to identify and extract entities from within a conversation. For example, dates, addresses, post codes, names, phone numbers, company names, even specific products. That’s pretty much all of the data an agent would gather to populate a CRM or similar system.

It’s not hard to see a further integration capability that would push those entities into the appropriate fields in the CRM to save the agent having to type and listen at the same time. This would free up the agent to be able to focus on the conversation and tune the inputs when it fails, rather than be distracted. The sound of tapping keys on a call doesn’t exactly make customers feel listened to, does it?

Creating new opportunities

CCAI Insights isn’t all about solving current pain points for customer service teams. It also enables completely new capabilities that businesses wouldn’t have even imagined were possible.

Custom highlighters

With Custom highlighters, you’re able to train the CCAI Insights NLU to listen for specific things within calls.

Let’s say you run a marketing campaign with a promotional offer, wouldn’t it be good to know how many people callers say they saw your ad?

Perhaps your competitor is doing something pretty well that you don’t know about. Wouldn’t it be nice to monitor every time a caller mentions a competitor?

How about analysing all of the calls customers make telling you they’d like to leave your service? Understanding the reasons why, at scale, is infinitely helpful.

CCAI Insights is making AI accessible

The possibilities with a solution like this are endless, and while it’s possible to build this yourself, Google through CCAI Insights, has just made utilising this technology a whole lot more accessible to all businesses, which could open up a world of opportunity for those who act.

    The world's most loved conversational AI event is back
    This is default text for notification bar
    Share via
    Copy link
    Powered by Social Snap