The Future of Customer Service, Part 4: The customer service agent of the future

20 Nov 2023

7.5 min read

This is a guest post written by Sham Aziz. With 20 plus years in the Customer Service game at companies such as Selfridges, Vertu, NET-A-PORTER & Ocado, Sham is a visionary Customer Service leader. This is part 4 of 4.

‘They look like me, but none of them are me’, I, Robot

In the first 3 tracks we discussed, reframing the role of Customer Service agents into a Personal Concierge. Someone who can step in and anticipate customer needs, moving from reactive, to proactive to predictive. How using CS as a competitive advantage, and removing friction can really improve the overall customer journey. We then went on to talk about true CSaaS, removing Customer Service as a function and focussing instead on the ‘skill’, enabling Brands to put customers at the heart of their business.

In this final track we will discuss putting agents at the heart of your business too. Helping customers with their needs, one does not exist without the other. 

Every year around budget setting time, we often hear the word productivity thrown around as a magic bullet to do more with less, or somehow hold to account those who could do more, but just aren’t. Heads and department managers within CS have had to find creative ways to make their budget stretch, and that matched with AI use cases in the contact centre, has oddly the opposite impact on productivity. If you automate the easy and repetitive stuff, you leave behind the trickier and meaty things for agents to deal with. The numbers will then of course show a drop in contacts per agent (CPA). 

Did we unknowingly suppress the need to ask the obvious question: will AI take my job or replace me? Not only did we forget about the elephant in the room, but it is now mechanised. How we got here is bigger than this mixtape but let’s take a step back to reality, by drawing the distinction between human-like vs. replacing humans, on behalf of the agents we work with. 

As AI improves, there will still be a gap between "human" and "human-like" responses. Great customer service requires true understanding. Imagine your wedding suit is delayed - would and could AI understand the importance and urgency, moving heaven and earth to find a solution? AI can sympathise to a degree but lacks human flexibility and autonomy. The customer service agent of the future is the ultimate problem solver, augmented by AI. When all else fails, the human agent swoops in to save the day, like a goalkeeper making an incredible last-second save. We've always seen agents as heroes, in the future this will still be the case.

Where are we now

Scripted responses – AI isn’t human enough and the scripted responses have the opposite but related impact of making humans more robot-like. Restricting the freedom that agents can have, leads to them naturally favouring process over judgement.

Lack of autonomy – Business rules are set that may prevent agents doing everything possible to help customers get what they need. Many agents fear making decisions that will ultimately help customers but which may overstep their authority and cause problems for them personally. 

Difficult to find information – what slows agents down is that the information they need is not easily available or easily searchable. Being able to get the context immediately, to read the case history is essential to provide the best response. 

Backlog pressure – agents are often targeted with answering a certain number of tickets per hour, even if they don’t necessarily resolve them. That’s because the backlog looks too imposing, and they’re always looking at the queue ahead. 

How do we get there

Fully autonomous agents

Too often we treat humans like machines. We set rules and boundaries and insist that they stick within them, and punish them for using their own judgement in a situation. That’s not a great environment to work in, so it’s no wonder there is high employee turnover in customer service.

But as the machines get better, the ability to program them to do tasks that only humans are currently able to do also gets better. Once a large portion of the repetitive tasks that humans have to do is automated, then businesses have a responsibility to think about how best to use humans.

Finding creative solutions to problems, seeing things that machines can’t, and truly understanding the human need behind a customer inquiry. This requires agents to have near full autonomy to solve the problem that the customer presents them with. Do whatever it takes to solve that problem. If brands can let go of some of the control and trust their agents to move between process and judgement, this will positively make customers happy. 

Augmented by AI

How would you explain your issue to AI? A superhero agent augmented with AI, could quickly see a summary of the case, sift through the salient points of interest, and get to the heart of the matter. This would negate the need for the customer to take a deep breath and explain their issue for the third time!

Then when the agent sets to work on solving the problem, the AI can step in to find the right knowledge base. A quick query can bring up everything that the agent needs to forge a solution. Any action that has a degree of complexity can be automated by AI, whether it’s generating a returns label, or booking a collection or any other such task like that. 

Augmenting allows agents to listen really well without distraction, and using AI to take over routine tasks through commands saves everyone time. Imagine no longer needing your teams to print out returns labels! 

Natural problem solvers

They will be the world’s best Googlers, the people who can find a solution to a problem and present it back with a “ta-da”, creating those “Wow” moments that really delight a customer and do wonders for the customer experience. 

The mindset of the customer service agent will change subtly. Removing repetition and ‘challenging’ them to find creative solutions makes for a far more interesting job. These problem solving skills would then have knock on effects to other teams and departments, as these agents move into other roles, and bring transferable super problem solving skills along with them.

By using good feedback loops into the business, each relevant department can use the data collected by CS to improve the customer journey. This joined up way of working to solve problems for the long term will make a massive difference to the customer’s life. 

Boundless creativity

AI image generators are very good at producing creative-looking images, but they still need a carefully constructed prompt to create something truly creative and personalised. If you remove the prompt, you end up with nothing.  

Take an AI powered stylist for example. Its algorithms might be trained on patterns it sees in fashion magazines, celebrity red carpets and perhaps choice Instagram accounts. This will most likely mean that a lot of its suggestions tend towards the samey, and may lack the finesse that a real stylist may have. 

For our problem-solving customer service agents of the future, they will have access to the best and most creative stylists and artists out there to tap into, and be able to find the right option for the customer who is asking. 

Imagine that one of the leading sneakerheads is a globetrotter from Sao Paulo, who really understands what is cool because she is setting the agenda. Being able to connect with her and get ideas would be invaluable to our agents, and that’s something that will be possible with modern technology. Custom curation from the best in the world, wherever they may be at that time.


While AI and automation will take on more of the routine tasks, the human agent will still have to solve the really complex problems that occur. Therefore the agent will need access to a wealth of information that can be easily searched and summarised. But they will also have to do the things that AIs will struggle to do, which is think outside the box. 

I’m hoping this rejection of AI as the side show to the main event of Customer Service superheroes, doesn’t come back to bite me, in a Skynet / Terminator movie plot scenario. Even with the amazing recall of some of the best memes on the internet, the gap between AI and agents is very real. 

Try asking ChatGPT for a joke. With lots of well structured prompting, it can probably find a good one from the languages that it is trained on, and maybe it could even explain why it’s funny. But could it truly possess the wit to be able to understand a situation and make light of it? The rate of progress may well render this conclusion useless in a few short months!

The Customer Service industry has focussed on control and processes that have slowly killed any creativity that will leave a customer delighted. We have pitted agents against each other in a race to be the most productive, which is not the same thing as being the most helpful. So let’s enable the environment for them to flourish and recognise their differences to unlock creativity and deliver happiness for customers. Because, whilst on paper Customer Service agents might all look like me, none of them are me.