All Aboard! Get Your AI Journey Off to the Best Start

The number of contact centres who don’t have an AI strategy has already dropped from 43% to 31% since 2023, highlighting that AI is becoming firmly embedded in the majority of CX strategies.

But what have the early adopters learnt so far and what can their experiences teach us about how to deploy AI successfully?

Our Editor, Megan Jones, spoke to Michelle Shaw, Martin Hill-Wilson, Alex Mead and Jeremy Watkin to find out…

Create Excitement About Your AI Journey

It takes time to change an organization’s behaviour, so don’t expect your AI solution to be welcomed with open arms! Instead, invest time and resources in engaging with your people.

Need inspiration? Here are some quick wins for creating positivity and engagement with your frontline – tried and tested by Michelle Shaw, (mostly recently) Innovations Principal at Capita Customer Management:

Share Teasers About the Specific Ways AI Will Help Your Agents

The timing of your communications is critical ahead of your AI deployment. We found that starting about a month in advance made the biggest impact.

Across this month, we did weekly teasers to get everyone excited, leading with “Wouldn’t it be great if you could do this…” to highlight the benefits that the solution would bring once it was deployed. It worked well!

For top suggestions on the potential uses for AI in customer service, read our article: Our Top Use Cases for AI in Customer Service

Encourage Agents to Record Videos of What They Love About AI

Want to harness the energy of enthusiastic early adopters or even those involved in a pilot? Then add video to the mix!

I got our agents on their mobile phones doing videos to share their positive experience with their peers. It didn’t cost anything! We just recorded them ourselves and put them on SharePoint for everyone to see.

Position AI as “Tech to Support Talent”

From a people perspective, your message at the start of your AI journey is crucial for winning hearts and minds.

A really positive way to achieve this is to refer to the new AI solution as “Tech to Support Talent” in all of your communications.

☆☆☆☆☆

Think Strategically

Lack of strategic thinking can create no end of problems when It comes to deploying a new AI-powered solution – from employee disillusionment to wasted budget.

Here are some of the most common mistakes to look out for when honing your AI strategy:

Stop Being “Too Busy” to Upskill and Educate Yourself

We’re still very much in an educational phase with AI, so decision-makers really should be upskilling themselves as much as possible, as Martin Hill-Wilson of Brainfood Consulting explains:

Martin Hill-Wilson, Brainfood Consulting — Martin Hill-Wilson

“For leaders, it’s about becoming more familiar with the language surrounding AI and the options available to them.

It comes down to understanding the art of the possible and how AI can support your operation in achieving its goals.

With such a disruptive technology in play, it’s critical to think it through properly and consider change management, the true value to your organization, and more.”

Never Assume Your AI Solution Will Be Compatible With Your Existing Technology

From an IT perspective, you need a firm understanding of what has already been deployed within your business. Not just that, but what’s coming down the line too!

Particularly in a large organization, you’ll need to bring together multiple roadmaps to make sure that what is on the AI deployment roadmap sits alongside your IT roadmap and other business plans for that year.

“There are some real horror stories out there from where this hasn’t been done thoroughly, including one organization I heard of that changed their telephony platform – a project that had been 12-18 months in the making – only to find that their new AI solution quickly became unusable, as the two solutions weren’t compatible.

The lesson here? Don’t make assumptions!” – Michelle Shaw

There are some other mistakes contact centres need to avoid when it comes to their technology. For advice, read our article: 3 Technology Mistakes That Are Probably Ruining Your Contact Centre

Write Your “AI Shopping List” and Be Clear on What You Need

A lot of people are lacking up-front clarity. Have you stopped to think about your contact centre’s specific needs and what you’re actually looking for?

“I’m very clear on this: you should have a shopping list ready for when you’re presented with different solutions, so you can confidently say “Well, I’m not in the market for that right now, but I’m definitely in the market for this…”

The biggest problem is that most people haven’t properly thought about what they want and why they need it. I would hazard a guess that only a few of the biggest brands are really managing their two-to-five-year outlook properly.” – Martin Hill-Wilson

So how do you decide what your next technology investment should be? Read this article to find out: What Tech Should You Buy Next? Here’s How to Choose

Don’t Try to Fix Everything, All at Once

A big mistake people often make is thinking you can just deploy AI, plug it in, and everything fixes itself. This is absolutely not the case, as Alex Mead, Global Customer Experience Director, explains:

A thumbnail image of Alex Mead — Alex Mead

“You should always start your AI journey with a proof-of-concept approach, because if you try a ‘Big Bang’ approach, you’ll just destroy your business.

For example, as most companies have an IVR, start by taking one option such as contract renewals and deploying AI there. Make sure you get that right first and then expand on it.”

Remember, it’s a marathon, not a sprint! Take a longer-term view and approach AI in small increments for overall success.

Be Transparent With Your Customers About the Use of Bots and AI

When deploying AI, be transparent with your customers about whether they are talking to a person or a robot, as hiding the truth can quickly backfire, as Jeremy Watkin, Director of Customer Experience and Support at NumberBarn, shares in his first-hand experience:

Jeremy Watkin, Director of Customer Experience and Support at NumberBarn — Jeremy Watkin

“My 10-year-old son recently had a poor experience with a live chat and eventually typed “You suck!!!!!!!”. After I called him out on his behaviour, he replied, “I didn’t know it was a person, I thought it was AI.”

For me, this highlights the argument about transparency, particularly as some bots even have human names.

Quite frankly, if you don’t make the distinction clear, you risk losing customer trust, as well as your people getting abused by customers if they mistake them for a robot.”

Make Sure Your Data Is Clean and “AI Ready” to Use

Contact centre leaders need to take a long hard look in the mirror at their true AI readiness and how well (if at all) their current data can be effectively utilized by any AI solution they deploy.

“All forms of AI rise and fall on the quality of two things – the quality of the model AND the quality of the data that feeds and trains that model. And guess what? Do we all qualify? Do we all have fabulous amounts of data – all clean, all available instantly?

No! Quite frankly, everybody is actually remarkably unprepared to do a good job on that. So, it’s important to be aware that this is a fundamental problem that, if left unaddressed, will make it difficult to get your ROI and make AI work in practice.” – Martin Hill-Wilson

Nip Scope Creep in the Bud

Beware scope creep!

“When you’re deploying new solutions, people tend to get really excited and quickly start to say “we could do this and this as well…” – all of which can throw off your budget and timeline for deployment.

To manage this up front, document the scope of the project and get it signed off and locked in by key decision-makers, so it doesn’t ‘grow arms and legs’ and get all confused before you even get started!” – Michelle Shaw

☆☆☆☆☆

Never “Take Your Eye off the Ball”

Another key area where AI projects can fail is when contact centre leaders shift their focus elsewhere too soon after deployment, resulting in loss of momentum and engagement, and an inevitable anticlimax for all involved.

Take a look at these quick tips to make sure this doesn’t happen to you…

Regularly Review Agent Feedback and Usage Levels

Don’t simply roll it out and walk away! Keep an eye on usage and remember to collect agents’ feedback too, as Michelle Shaw explains:

“I used to drop post-implementation surveys at around 3 months in, asking all agents ‘What do you like?’ and ‘What do you want to see more of?’. I would also keep a close eye on uptake and who was actually logging in to use the new system.”

If you are looking to improve how you gather staff feedback, read our article: 29 Ways to Transform Your Call Centre Staff Surveys

Don’t Deflect Everything to a Bot and Ignore What Happens Afterwards

Monitor metrics and look at the actual conversations your customers are having with your AI bot. If customers are swearing at the bot or constantly getting stuck, you really need to know so you can improve the bot’s flow and make it less complex.

“A mistake contact centres often make is assuming their customers are having a great experience with the bot without closely monitoring the situation. It’s important not to deflect responsibility to the bot and ignore whatever’s happening. There’s always a lot to be learned along the way!

The goal should be to reduce the instances where customers need to contact support in the first place. To do this, you should always be reviewing the conversations customers are having with the bot to ensure it is solving their issues.

Initially, we made the mistake of assuming our bot was resolving customer issues when, in fact, it wasn’t. Now, even when a customer says “yes, my issue is resolved,” we still keep the ticket open and have an agent review it before we close it off. Having humans involved in the process is crucial!” – Jeremy Watkin

You also need to make sure that any bot you deploy has the right fail-safes in place to avoid unwanted interactions. For advice on this, read our article: F*** This! How to Make Sure Your Chatbots Don’t Swear at Customers

Don’t Let the Negative “Water Cooler” Conversations Win Out

Do you find there’s people in your team, including managers, who become so focused on what is going wrong, they ignore the majority of situations where things are going well? It’s the same with AI!

Yes, there will be teething problems along the way (and they will inevitably feature in your daily team chit-chat), but it’s important to proactively offset these negative conversations.

So, make time to nip any negatives in the bud and focus on regular, positive contact centre-wide communications that celebrate what’s going well.

Make Time to Consider the Wider, Transformational Impact of AI

Don’t just focus your attention on the short-term, widely available quick wins! Think about the longer-term transformational impact of AI too!

“AI now has the capability to understand natural language incredibly well. For example, imagine a customer asking an airline, “Why has my flight been delayed for two hours?” or “What options do you have for airport parking?” They should be able to speak or type that into a natural language engine and let AI handle the rest.

That’s where AI should be. Once you get that right, you can focus on everything else. But I still don’t hear people thinking that way. It’s the direction AI needs to move in, but it isn’t widely in use yet. Everyone else seems to be focusing on Gen AI to support agents with chat sessions or summarizing cases. But I believe we should focus on the front end first.” – Alex Mead

Double-Check Everyone Has the Kit They Need to Be Able to Use the New Solution

And finally, don’t make assumptions about the technology and kit your agents actually have in front of them – especially those working from home!

For example, if agents need 2 monitors to enjoy the new system, do a survey and check who has the right kit. It could be the real reason they’re not using the system like you’d expect.

★★★★★

Contact Centres Are Still at the Beginning of Their AI Journey

Wherever you are in your AI journey right now, we’re all still very much at the beginning of understanding exactly how and where this newest technology will make a difference to our personal and professional lives.

What’s clear, however, is just how much a clear, strategic vision, which engages with agents, customers, and IT teams alike, is critical to getting your AI journey off to the best possible start.