Analysing contact centres
Contact centres are home to some of the richest data a company may have. Now, as Heather Wright discovers, the move is on to fully utilise that data through efficient analytics.
No longer just the ears and voice of a company, contact centres are using increasing numbers of communications methods, gaining huge volumes of data in the process.
If last year was about the push towards including social media in the contact centre, this year may well be the year of analytics.
Zeacom senior contact centre specialist Tom Farquhar says more and more analytics is being seen in the contact centre and there's an increasing demand to analyse unstructured data – analysing recordings for both content and sentiment. With costs coming down and ease-of-use and interfaces improving, plus organisations' desire for more analytical knowledge, he says speech analytics is a rapidly growing area with easier justification for the costs.
While traditional dashboard metrics, such as calls per hour, average handle times or even first call resolution, focus retrospectively on agent performance, more valuable insights can come from more in-depth analysis.
Traditional call monitoring – typically of five or 10 calls per agent per month – is a small sample of calls, Farquhar says. Speech analytics extends that and can listen to all of the calls, searching for key phrases, such as 'complaint' or 'refund', to provide a more comprehensive, and objective, view of why customers are calling.
“If you're relying on agents to provide wrap-up codes, marking up why people are calling, it can be limiting and it is very subjective. Speech analytics is much more accurate and objective.”
Stress levels in agents' voices can also be detected and flagged, helping contact centre operators better understand what is causing the stress.
Farquhar notes that the phrase 'I want to speak to a supervisor' is never welcome news, and agents are taught to try to remedy the situation rather than passing the call on. “Speech analytics enables you analyse all calls [where that comment is made, whether it is escalated to the supervisor or not], and understand what happened to make a customer want to escalate it to a supervisor.”
Staff skill sets can also be improved. “Sometimes people don't know they're not doing well in some areas. And this can help highlight those areas so they can be improved.”
A report from Zendesk, released in May and based on customer service and support interactions between 16,000 participating organisations and their customers across 125 countries, found that organisations which take advantage of analytics reporting provide faster customer service with 12% shorter first time reply. Enquiries were also resolved 16% faster, giving truth to Peter Drucker's 'What's measured, improves' quote.
More information, however, is both a blessing and a curse, and Farquhar warns that contact centres need to understand their KPIs, the keywords and phrases they need to be listening for, and what they're trying to get out of speech analytics, before embarking on a new analytics programme.
More than that, speech analytics calls for dedicated resources. “If you don't have a dedicated analyst, you'll be putting in something that you're really not using.
“It's really critical before you even consider implementation to ask yourself if you are effectively reviewing and using the information you already have. If not, you may not be ready for speech analytics.”
Dan Scheltinga, Zendesk's Asia Pacific support and services director, agrees that analytics is big business and enabling companies to make meaningful decisions. But other trends are also changing contact centres.
“We're seeing an increase in internal use cases for for contact centres. Traditionally the focus has been on external customers, but increasingly organisations are using a customer services solution to service internal customers for purposes like an internal IT department, HR or facilities management.”
The shift to self service is also continuing, Scheltinga notes. “There's a massive uptake in knowledge bases and we know that given the option to self-serve most will try that first.
“We're seeing knowledge bases in two key areas: ones that the organisation itself popultes with information they think customers will find useful, and one that is a Q&A, where customers can help other customers.”
With advanced ticketing systems customers can log requests and receive auto-suggestions based on metrics, and links taking them to knowledge base articles, again funneling them through the self-service model.
Increasing integration – with third party apps within an organisation, or even with other companies, is also driving change, he says.
“There's a thirst for integration in the market at the moment. You could have a shared service with other partners, so you can share tickets with, say a back-end voice provider. Gone are the days of having to draft an email to the other provider about a customer issue. The two companies can easily work together to satisfy their customers.”
Internally, there's demand for integration which empowers the contact centre agent, providing them with access to multiple sources of information, without having to jump across applications.”
Meanwhile, the drive for multichannel continues with customers wanting to contact organisations on the medium they most prefer. SMS queueing, is something being seen more in New Zealand than the US, says Farquhar, but he expects it to only get bigger in the future thanks to text happy younger generations.