How Contact Centre Conversations Became Broken, and How to Fix Them

Jimmy is a self-confessed failed actor, failed script writer, and failed director. But it’s not all bad news, because he’s taken these failings and his performing arts experience and applied it to contact centres to expose 3 myths that your business can address to impress customers, retain agents and ultimately improve revenue.

Myth 1: Removing scripts empowers performance

Roughly 10 years ago contact centres everywhere decided to move away from scripts and teach agents to improvise in the hopes of empowering them to deliver a more human customer experience. Yet even for trained actors who spend day after day performing, improvisation is exhausting. 99% of all performances are scripted, and even some of the best actors in the world struggle to improvise for long periods of time – so how can we expect our contact centre agents to be able to do this?

The real key to improvement is to teach agents how to improvise around the edges of their scripts to add a bit of colour to the conversation – it shouldn’t be the main body of what they’re doing. So, how can contact centres practically apply this? You should be using conversational guidance software to intelligently guide your agents based on customer data – resulting in the best experience, best journey, best recommendation and most importantly the best outcome. In other words, the Next Best Action.

Myth 2: Speech analytics tells you what’s going on in your calls  

Speech analytics has been misused for decades – any one of your agents can tell you what’s going on in your calls, but you should utilise speech analytics to tell you what’s actually working, and why. Use this vital tool to understand how closely your agents are delivering your ideal customer experience.  

The key benefit of speech analytics is that it can help you to recognise which parts of the script drives positive and negative outcomes – what was the reaction of your audience? It’s a bit like actor fluffing their lines and being told how to continue by the director on set – use speech analytics to monitor, prompt and improve over time.  

Speech analytics can also be used to automatically identify complex characteristics in customers and alert the agent (usually, this is more accurate than relying on agents who are busy multi-tasking and often under time pressure). For example, a customer who seemingly struggles to comprehend what is being asked of them is very likely to be classed as vulnerable. If a vulnerable customer is dealt with in a poor manner, then speech analytics technology doesn’t just detect this but can also send an email alert to ensure the mistake doesn’t happen again, leading to continuous improvement for agents and better outcomes for customers.  

Myth 3: Data should tell a story

Enough about the actors - what makes a good director? When they are on set, they need all the best tools they can get to enhance the script and performance. They should be communicating with precision, offering clear direction and filming as many takes as necessary to get the best performance. We must continuously highlight the actions within data points that need to be improved upon.

Controversial statement incoming: data visualisation is dead and should not be used for operational tasks. It is imprecise, and it adds layers of complexity that requires time to interpret – time that contact centre agents just don’t have. The fact it is open to interpretation in the first place is an issue in itself. Feedback should be straight forward and actionable. So, how do we achieve that?

Put yourself in the shoes of a contact centre agent, look at the two images below, and ask yourself, ‘Which is easier to understand, and which would I be more likely to take action from?’:

Even with simple scatter graphs such as the one shown above, people can hide behind the fact that it both takes time to understand, and that it is up for interpretation. There is no room for confusion in the email - it gives simple calls to action that anyone can understand, backed by actual, significant data. It saves time by using precision in directing you to who to speak to, and as a bonus it’s measurable and repeatable. Through pixel tracking, we know who opened the email, and we know who clicked through and scheduled a coaching session.  

The most beneficial thing a contact centre can do is to combine these technologies to drive continuous improvement all together. We’ll show you how this pans out in a super quick insurance retention case study:

Phase One: Act. The client introduced new scripts for contact centre agents, but in the first 1-4 weeks saw a 1% improvement in outcomes (worth roughly £800k).

Phase Two: Listen. Through the use of speech analytics, we discovered only 50% of the agents were adhering to the new script!

Phase Three: Learn. We flagged low script adherence and created a coaching feedback loop, which drove adherence up to 70% within weeks 4-8.  

Result: In just 8 weeks we found a 2% improvement in customer outcomes – worth over £1.6m. This demonstrates the importance of Next Best Action, Speech Analytics and Coaching working in tandem.  

​People often fall into the trap of believing that implementing AI and data automation technologies in contact centres is alien and difficult. But in steering away from them we put our customer contact at a disadvantage. Do chat bots have scripts? Yes. Do we monitor and modulate chat bot performance to get better outcomes? Yes. Do we give them clear logical data points instead of visualised data? Yes. In many ways we give robots better information than we’re giving to our agents, so it’s time to fix that and give our real talent the upper hand.