A large UK based Telco
Transform contact centre compliance from a low-value obligation to a source of value-added customer insight.
In 2017, a large telco came to us with a request - “Help us automate compliance monitoring across more of our customer calls” . Through our method and process at the Applied Innovation Exchange, we not only achieved this, but we went above and beyond to create a working proof of concept using the application of emerging technology. I was part of an ambitious project to help call centres gather better insights to their calls and improve the overall experience for customers, whilst achieving compliance requirements set by both the client and Ofcom - the regulator and competition authority for the UK communications industries.
Our goal for the project was to design and build a solution to automate compliance monitoring and deliver insights in, or near to, real-time, by applying the use of emerging technology. If we could create something with extra value: bonus points.
A platform that uses speech to text technology to analyse calls and ensure compliance is being achieved. The platform also uses sentiment analysis and other characteristics to provide an agent scorecard to identify any areas for improvements in the agents role
The client currently employs an offshore team to monitor their call centre staff and ensure that they are meeting compliance regulations. The current process is manual, time consuming, and very subjective depending on the person scoring.
Due to the sheer volume of calls, only 2% are monitored at present. 11% of which fail compliance, and costs the business 98p per call. The client also engage an analytics software which can pull key insights from calls, however team leaders have to manually request an investigation into specific calls, and it can take up to 2 weeks to receive the insights back.
At the outset of the project we had lots of information and data from an operational and business level, however not much in terms of what was happening on the ground in the call centres. Without any solid pre-existing insights, I partnered with our product owner and we went on a road trip to the north of England, visiting 2 call centres to speak to the agents. We also went to our client’s head office to interview senior stakeholders from a business value perspective to validate the information we had so far. We asked agents about what frustrations they have in their roles, and tried to identify why calls might not meet compliance requirements if the first instance. We also listened in to some calls to really put ourselves in their shoes. We then spoke to the team leaders who gave us some real insights into what would be valuable for them in a potential solution. Finally, we chatted to the call centre operations leads, to provide us with insight from a call centre performance perspective
Customer completes automated identity and verification in their own time, enhancing the user experience by allowing them to gather their documents/passcodes in their own time.
100% compliance requirements are met.
Sentiment analysis detects various different audio features, such as pitch and tone of voice, good indicator of the level of rapport building.
We added a word cloud at the call level and also at the campaign overview level. This quickly gives anyone investigating a visual summary of the types of topics trending on the phone lines.
Each call is recorded giving all users the opportunity to listen back to any call. Easily filtered by time, date and agent. Could also be integrated with client CMS. Allows team leaders to effectively coach agents.
We included a scoring feature which analyses the overall performance of an agent measure against the compliance metrics. This element of gamification would drive more engagement for agents and reinforce the gamification that already exists. This could be screened in the call centre, along with key call stats that are already displayed.
To add a gamification element in the solution, we included a leaderboard so that agents and team leaders can see who is performing well. This could engage the team in friendly competition to compete for those sought after rewards.
Speech to text
Speech to text give an accurate transcript of the call which could be forwarded if required.
I validated the design with a small subset of agents, who agreed that it was straight forward to use, and would be beneficial for them to have access to their individual call libraries - something they do not currently have access to. They also liked the gamification aspect. This gave us confidence that our MVP was something that woyld prove valuable. Obviousky if the solution progressed further this wold have been subject to more rigorous user testing.
We demoed our solution to our client’s senior executive teams and told the story of how we got there.
Although the solution was not taken further in this shape and form, the client were very impressed and the engagement led to further investment in another area of work.
The technical platform has now been developed further and is due to launch as an anonymised version, where users can gain valuable insights into a vast data set of calls. This is due to be launched in 2019.