There are some big questions that we shouldn’t shy away from. People are asking if AI will replace their jobs and I cringe when I hear people answer, "No, of course not, this will never happen”. It is highly likely that at some point in the future, AI will replace a significant number of human tasks. But the key word here is tasks, not jobs. AI is not going to replace jobs overnight, but the removal or enhancement of many tasks involved in a job is already here. Like many organizations, AI is an incredible opportunity, but also a huge responsibility for Probe Group as a company to ensure we are equipping our people with information about what’s happening and what it means for their roles. The AI transformation is underway and while it won’t be instant – and there are a lot of positives – there’s no doubt that it’s daunting.
Probe Group is a naturally digital, uniquely human organization that places people and culture at the heart of everything we do. In a lot of ways, I see what we’re facing with integrating more AI into our processes as typifying the dilemmas facing all kinds of industries. It’s such a fast-changing landscape and it’s our job to make sure that it is positive for our people and our clients. Here’s a little of how I see the opportunities and the challenges playing out in the near term.
Bringing together people and digital
Delivering leading experiences requires the integration of people and digital technology. We believe real impact comes when these two elements are combined rather than kept separate. The products and capabilities we are developing bring digital and AI technology together in a naturally human way. We work with our strategic partners to leverage their digital capabilities, enhanced by our specific operational expertise and IP, to deliver exceptional customer experiences.
Probe Group’s strength and differentiation is the way we combine human and digital services, together with our local Australian experience to serve clients all over the world. Our value lies in offering clients complete solutions that incorporate the people, skills and resources needed, with leading technology products and platforms.
We see it as a symbiotic relationship – we can help our people deliver great service for our clients (and their customers) through adding relevant digital technologies to help them perform their role better and conversely, technology solutions alone, without highly skilled, qualified people, are not as effective.
It’s a very important interdependency and what I like about it is it represents elements of the future way of work. The future is people and machines working together as a ‘blended workforce’ and that’s really what we offer. Probe Group brings together the best people with the best technology.
AI in action for our people
Our view of AI and I’d say it’s shared with many other companies is to be a companion or assistant that enables a person to perform their role at a higher level.
We are deploying AI technologies to create experiences that customers love, that our employees love and our clients love. The first is around the customer experience. An example is when we deploy AI technologies to help our client’s customers without them needing to interact directly. We deliver this using conversational AI voice and chat agents with natural language capabilities to help customers resolve their enquiries without the need to speak to an agent.
As well as focusing on always improving experiences for customers who come to us, we’re also working on using AI to architect experiences that our customer service consultants or representatives love. That starts with taking away mundane tasks they don’t enjoy to give them more time to create a better connection, have a higher empathy in the interaction and provide a better, more human conversation. Our goal is to balance the efficiencies needed to run effective contact centers with the uplifting quality of every human contact.
This includes benefits such as real-time transcriptions and summaries of live conversations. The notes at the end of the interaction are automatically summarized and sent to a CRM system. We’re also excited about the potential of AI companions and assistants. These AI assistants are able to listen to live conversations and proactively suggest contextual knowledge via pop-up recommendations as well as provide additional support to the consultant and make their job easier, particularly if they’re dealing with an unusual enquiry.
The third category of how we are using AI, is building manager assistants. We know that one of our managers' key responsibilities is supporting their teams through better coaching. You can vastly improve the quality and tailoring of that coaching when you have more context, rather than a blanket approach. AI enables us to do quality assurance (or QA) on 100% of interactions as opposed to the industry average of less than 5%.
We can use AI technologies to scan all the call summaries and identify if there are issues or concerns. That might be sending insights directly to our clients to let them know that a particular part of their organization is having issues, for example for retail clients we are able to identify trending issues such as deliveries from specific stores or distribution centers, and provide that insight to the Operations team to address the root cause of the problem. Equally, it might be how we’re handling a particular type of call in our center. We can pull this out and suggest to the manager that they have a coaching session with the agents who are struggling with handling a particular request or situation.
Having AI tools scan those summaries shortens the loop to bring much faster, contextual and more relevant insights.
Better questions are the secret to high-performing AI
Generative AI will require a balance of discipline and creativity if we are to get the most out of it. The implication is that we are beginning to see new roles emerging, such as Prompt Engineers. Given answers are now far easier to obtain than ever before, the way you develop and ask questions has never been more important. If you don’t ask a good question and understand how to structure a prompt, you are unlikely to get a quality answer. The skill of asking good questions and engineering the right prompt will be in high demand.
As it is, I don’t think organizations devote enough time to think about how we ask questions. AI is the world’s first true socio-technology - machines to try to mimic what humans do. It is trained on data created by us, observes our behaviours and learns from how we react. That’s why social skills are going to be so in demand to deploy this technology.
Curiosity is so important and yet many businesses haven’t got a discipline around asking questions. They’ll ask “Why are my sales down this week?”, but it’s reactive. Instead, the approach should be to set out in a plan that outlines the knowledge they want to gain over the next 12 months and these are the questions they need answered to do it, rather than scrambling for information ad hoc.
We need to think about approaching knowledge by building a sequence, writing down a series of questions to gain deeper knowledge and asking further questions on the back of the answers.
We also need to think about how we connect the models we build. For example, you might build a model on customer churn, a separate model around offer responses and another on lifetime value. I would say 99 times out of 100 organizations build these models as discrete projects. They’re never linked together. Those models represent three questions:
- Is this customer going to leave?
- Will they take this offer up?
- What’s the impact on lifetime value?
The outcome of one of the models - aka the answer to the question - should inform the input to the next model, aka the next question. But it’s not architected like that. Organizations spend a staggering amount of money on data and analytics without a roadmap that links these models together. The secret to unlocking generative AI is the ability to ask good questions. That’s why there’s so much emphasis on the job of the prompt engineer.
High-performing companies understand the questions they should be asking and are working very hard to answer those questions. Not only is it far more efficient when you build these capabilities, but it is also obviously more impactful.
Answering these questions – before they’re asked
Coming back to CX in contact centers, I’ve set our data science team one goal: to predict every single call/interaction that comes into a Probe Group service client and why that interaction has occurred, before it occurs.
The team quite likes the challenge - it’s like an XPRIZE, can you do this? It will never be exact, but if we can get close to answering that question the value for our whole business and our clients will be incredible. It will flow into resourcing, workforce management staffing, being proactive with solving a need before the volume comes in, and helping improve core metrics like Average Handling Time (AHT) and Quality of Service (QoS). Before this wouldn’t be possible but with AI, the goal is within reach.
We support a large energy distributor in Victoria, who had previously had a big storm event where more than 200,000 customers lost power, with some outages lasting weeks. Despite working long shifts, their contact center team struggled to manage the volume of calls.
After that event, they approached Probe Group to design and implement a system that could cope with such an occurrence, which duly struck just a few months later, impacting more than 250,000 properties. At the peak of this storm, there were 600 calls per minute and despite not being resourced to handle such an extraordinary event, Probe Group were able to scale to meet the demand using a combination of our proprietary Oration AI technology and our flexible staffing model.
While that’s a good story for that client, there’s a tremendous opportunity to use AI to do even better. We can use weather data and be proactive in anticipating the impact of the storm. We can message the energy provider’s customers directly ahead of time and make sure that we’ve got banners and messages in different channels before the storm even hits.
CX is about always listening - responsibly
The fundamentals of CX are always the same: we need to listen to customers and respond in a contextual, relevant way. A deep understanding of customers combined with behavioral insights and data is the absolute foundation of AI-augmented CX. You have to listen to and measure all interactions across all channels. Once you capture the information, you can build the AI that helps you better understand customers’ needs and lets you respond in a relevant way.
This by the way isn’t all about selling. We’re working on a better understanding of community resilience, vulnerability and hardship. Our objective is to help our agents, and our digital channels, be more empathetic by using technology and AI to improve our ability to identify sensitive needs and respond with the appropriate care, consideration and privacy.
Probe Group recently released our Responsible AI Policy. They are non-negotiable commitments we have made to our people, our clients and the customers we interact with. Everything begins and ends with these commitments and we’re confident that within that framework we can use data and AI responsibly to improve the customer experience and our service by being more contextual, more relevant and showing higher empathy.
What I’m most excited about in all of this is the scope of possibilities we’re exploring. When new tech comes out, it begins with narrow use cases in the way it’s applied. That’s certainly been the first wave of this era of AI. When you integrate AI into your processes, it will change the way that people work. AI will go from solving one task, to two tasks, to five and take on increasingly complex tasks until it’s now performing a part of the role. That is going to open up a whole new level of customer service. It allows our people to be more human when engaging with customers.
As we become better at using the tech, we will get the most out of our human capacity. Without wanting to sound grandiose, I genuinely see this as significant for society and for the jobs our people perform – we’re already seeing examples of it playing out in discrete ways. When we put it all together in a responsible and controlled way – and I can’t overstate the importance of that caveat – that’s when we will really benefit from AI-enabled transformation.
