Call centre quality assurance is a critical component for most call centres, whether regulated or not.
But it’s also a time consuming, costly, resource intensive process, manually listening to calls from agents and marking scores on a scorecard system. It’s a process that can easily take 30-40 minutes for a 10-minute call.
Even with the best of intent, most general function call centres audit only around one percent of all their calls, leaving 99 percent of calls unmonitored – and potentially exposing your organisation to risk, including poor customer experiences.
Specialised contact centres which are heavily regulated may be auditing at a costly expense 25-80 percent of calls, or even 100 percent depending on requirements, but they’re the exception to the rule.
There’s also another flaw inherent in the process.
Often call centre workers are required to follow a script on which they’ll be marked. Fail to follow the script and that impacts their KPI and performance review. But in following the script conversations may be stilted or unnatural, with agents asking random questions which don’t fit into the flow of conversation.
Consider too, the data gold mine that is your call centre, with potentially millions of calls sitting in in your file repository. That’s data that could potentially be mined and provide a wealth of insights for your business.
What businesses need in the modern world is a cost-effective, scalable and reliable solution that can not only QA 100 percent of call, but also derive near real-time information from calls to inform the business so they can make rapid decisions – whether strategy, marketing or customer service.
And that’s where QBOT® (Quality Bot) comes in. It’s an AI-powered compliance and quality assurance developed by Quanton, leveraging a locally developed AI algorithm to accurately assess calls based on an organisation’s QA and compliance regime.
100 percent auditing, with a 50-70 percent reduction in time and cost
The solution we have developed provides a near real-time dashboard for QA and the contact centre performance, which as well as enabling you to audit 100 percent of calls, can detect customer and agent sentiments and create triggers around potential issues, while presenting insights mined from those calls so you can understand your customers better – and build new efficiencies into your business.
QBOT, which is already in use at a number of customers, including Momentum Life, works with existing call centre platforms such as Genesys and AWS, so you can sweat those platforms.
Calls are saved in the call recording repository. Our AI and machine learning solution then turns the audio into text and analyses calls, putting them into an intuitive dashboard to provide the insights for your business.
Via a simple to use portal, and in as little as a minute, you can get a good idea of what the call was about – the reason for it, how it was resolve and intimate analysis. Call summaries and the AI scorecard viewed on the same page, along with the transcript and audio files.
The AI will also provide a self-assessment of its confidence on the call transcript, and we’ve seen great results for heavy accents. Different language models can also be used – some organisations for example have Korean Chinese call centres and our system can transcribe and translate calls into English for QA and analysis.
Mining for insights
All the information you’re collecting via your call centre can provide valuable insights for improving your business.
Consider the example of a customer calling in because their broadband isn’t working. They’re assisted to replug the modem and the issue is resolved.
Those insights can be used to reduce customer calls through providing them with that advice – say through a chatbot – before they call in, so only the more complex issues get sent through to your call centre.
One issue a lot of call centres have is categorising calls, with large numbers of calls often ending up in the ‘other’ or ‘general’ category. That might be because agents are having challenges selecting between the different categories.
We can train the AI to identify the team a customer needs to be talking to – whether customer service, technical support, sales or onboarding, based on existing call centre categories.
That simple step makes a huge different to the call centre in understanding the drivers of why customers are calling and putting in measure to reduce call numbers.
AI can also be used to handle call summaries, so agents don’t have to write up notes, and ensuring consistency, and to mine post-call action items. We can add a trigger for action items to trigger an automated workflow or automation in your CRM system, for example, to automatically send out follow-up documents or text message, reducing the time your agents have to spend doing those processes and freeing them up to help customers.
Resourcing wins
With the near real-time view of call centre operations, call centre managers can quickly identify where resources are most needed and divert resources there as needed, leading to better workload management.
Opportunities for coaching or training of employees can also be quickly identified and if an employee is struggling on a particular day, call centre managers can quickly step in to provide assistance and support.
For the contact centre, QBOT is all about enabling you to run quality assurance across 100 percent of your calls, while getting near real-time sentiment and performance results, identifying coaching opportunities and knowing your customers and the drivers of the calls.
For managers, it enables use of customer insights to steer business decisions and strategies, the building of a culture of continuous improvement and mitigation of regulatory and compliance risk.
And there are benefits for sales, marketing and R&D too. Coupled with closing more deals through understanding key customer behaviour, and insight driven product and service development.
And it’s all here and now – a real life AI solution for your business. If you’d like to see how Momentum Life put it into practice, you can read the case study here.
