How AI is Redefining Customer Support: A Global Leader’s Perspective

In October, I had the privilege of speaking at Intercom’s Pioneer Summit in London to discuss how Lightspeed is leveraging AI within our customer support group.

As someone who has been part of the customer service and technology space for nearly 18 years, I, and other leaders globally, recognise the potential benefits of automation and AI to enhance customer support through self-led help coupled with a high-quality, personal human touch—at scale.

Off the back of this talk, I wanted to dive a little deeper into how Lightspeed utilises AI to better support our internal teams and our customers and my predictions for how AI will evolve in the coming years.

Lightspeed and AI

As a tech company, it’s a given that Lightspeed should be exploring and implementing AI in different ways, and as a business, we’ve been very forward-thinking in the AI space.

From utilising AI in our internal operations and releasing customer-centric AI features within our products, we are consistently seeking to understand how AI can enhance our own business operations and those of our customers. 

With several AI initiatives already underway, the next logical step, for me at least, was to introduce AI into our customer support processes.

Leveraging AI to enhance customer support 

For the last 18 months, we have been leveraging AI to assist in the resolution of new support enquiries. We have achieved this through a number of initiatives which utilise our comprehensive customer help centre and the collective knowledge of our amazing support team members.

We’re using AI in conjunction with our support team.”

By implementing AI into our customer support workflows, we’ve been able to power an AI support assistant within our online chat experience directly within our Back Office and point of sale solutions. This allows customers to ask questions 24/7 and get immediate assistance, with our friendly team still available in case more help is needed.

I think this is an important point to highlight—we’re using AI in conjunction with our support team. We implemented Fin AI to help our support teams, optimise workflows, and provide faster resolutions for our customers. However, if the customer can’t find the answer they’re looking for, our team is only a few clicks away to dig deeper into any queries. 

We also use AI to power our internal help tools to provide our team with the best possible data and information when solving customer challenges. Thanks to these new workflows, we have been able to speed up response times and become more effective at solving issues, which are huge wins for our team and customers.

Not working on a hypothesis

While we’ve had great internal support for this initiative and witnessed some great success, I understand it’s often difficult to pitch an idea, particularly a new concept like AI, and get full backing to implement it. 

One question you often hear when discussing the implementation of AI is: how do you get stakeholder buy-in?

Change management isn’t always a smooth process, and there will naturally be some uncertainty when implementing new technology or processes. The best way to manage this is to test and learn to demonstrate your proof of concept.

Fortunately, in my experience, it was easy to present a business case to use AI because we weren’t merely working on a hypothesis. The Intercom platform allowed us to test, learn and tweak different scenarios, making it easier to deliver a robust proof of concept. We were very clearly able to articulate why we were doing this, what results we wanted to see and what the business could expect.

I think this really helped our whole journey to approval. 

Timing is everything

Another common question I get about adopting AI into our customer support workflows is: how have customers responded?

For me, timing is everything, and we implemented AI at a time when AI has become somewhat commonplace in our everyday lives. Things like ChatGPT have become pretty prominent, so the concept of using AI is no longer abnormal. I think this has helped in our rollout, as customers are more willing to interact with AI than they would have been 18 months ago. 

Introducing AI into our customer support has added an additional layer of convenience for our customers as they can ask questions whenever they want, and of course, our friendly team is still there, one click away—so it’s been really well received. 

The challenges of adopting AI

I touched on this earlier, but change management was one of the main challenges we experienced when implementing AI. While we were all on board with rolling out AI, it wasn’t as simple as flipping a switch overnight. 

With hundreds of customer support agents spanning multiple regions and languages, we needed a clear, systematic approach when introducing our new AI workflows to ensure adoption and go-live were as smooth as possible. 

Setting up for success

To ensure the team was set up for success from the get-go and to minimise any disruption for our customers, we focused on a few key areas to make sure everyone was equipped to succeed.

1. Training

The initial area we wanted to focus on was training. In the weeks leading up to the rollout, we conducted several training sessions to ensure our team was fully prepared. As a first step, we utilised Intercom Academy to allow our team to learn more about the platform and give them a firm grasp of the tools. In addition to this, our in-house training team created customised training materials tailored to Lightspeed’s unique processes and workflows.

2. Ongoing support and enablement

Following the initial training, we adopted a “hypercare model” for post-launch support, which included dedicated Slack channels and forums where team members could ask real-time questions or report any issues. This approach enabled us to resolve concerns promptly and make any necessary adjustments, ensuring a smooth and successful rollout.

3. Clear communication

Aside from training, it was also essential to ensure that we effectively communicated the upcoming changes and new workflows to other teams that might be impacted by the rollout. And it was important to do so well before any changes went live. By giving ample notice and providing clear instructions on what to expect, we were able to minimise any surprises on launch day.

4. Align vision and goals

We recognised that managing change within a large, geographically dispersed team required a well-coordinated approach. To ensure everyone was on the same page, we worked closely with our leadership team to align on the goals and benefits the AI implementation would bring to the business. By engaging key stakeholders early in the process, we were able to build consensus and foster a sense of ownership throughout the organisation.

The benefits of adopting AI 

We’ve seen some incredible results since adopting AI into our customer support workflows, from reducing training time to resolving more customer queries. Here are two key statistics I’d like to call out.

• More than 65% of new conversations started with AI do not need to be transferred to our team.
• Our Internal AI Co-Pilot tooling is allowing our team members to operate up to 31% more effectively in resolving issues on first contact.

Another benefit we’ve realised is being able to better utilise the wealth of customer-facing content that we’ve created and maintained over the years. AI has helped us to create a simple way for people to engage with our help centre content, selecting the most relevant content based on the questions presented. 

Team development as a result of AI

One of the biggest benefits of implementing AI into our support workflows is the amount of bandwidth that’s been unlocked. This has given us more time to look at what we do next, such as exploring the possibility of further developing new functionality to optimise our workflows. 

On the human side, we’re now looking at what the role of a support specialist looks like now we’ve augmented what questions are being fielded by our team members and what’s being tackled by AI. We can take stock and go, ok, what can that person now do for their own personal development that benefits them and our customers.

What does the future look like?

While it’s always tricky to predict what’s going to come next, one of the big advances I see happening in the AI space is around data and analytics.

Here are a few ways in which I think AI will advance in the near future.

1. Reduce reliance on human categorisation 

AI will increasingly automate the classification and organisation of data, reducing the need for manual intervention and minimising the risk of human error and bias. This shift will free up valuable time to spend inspecting these insights and making efficiency-boosting decisions.

2. Add depth to data and insights 

As AI models become more advanced, they will not only analyse surface-level metrics but also uncover deeper, more nuanced insights. AI can identify hidden correlations, emerging trends, and complex patterns within data that traditional methods might miss. This ability to provide more granular and meaningful insights will allow businesses to make more informed, data-driven decisions, unlocking new growth opportunities.

3. Better detect opportunities for improvement 

AI-powered analytics will excel at pinpointing inefficiencies and areas for improvement across different processes and customer interactions. By continuously analysing large volumes of data in real-time, AI can detect anomalies, predict future trends, and suggest actionable optimisations faster and more accurately than human analysts. This will lead to quicker decision-making and more agile operations.

4. Enable large-scale analysis quickly and easily

AI will significantly enhance the ability to conduct large-scale data analysis quickly and easily. It will handle vast datasets in real-time, allowing for the rapid generation of reports and dashboards. Complex analytics that once required significant time and resources can now be performed efficiently, enabling businesses to keep pace with market changes and customer demands.

5. Fuel stronger collaboration between customer service and product teams

AI will bridge the gap between customer service and product development teams by capturing detailed customer feedback and sentiment at scale. AI tools will analyse customer interactions, reviews, and feedback, offering precise insights into user preferences and pain points. This will enable customer service teams to communicate actionable insights to product teams, fostering better collaboration and ensuring product updates are driven by customer needs. Ultimately, this will result in more aligned and user-focused product development.

There are so many possibilities for AI and data analysis, and I’m excited to see what changes happen in the coming years. 

In reality, we’ve only just started to scratch the surface of artificial intelligence, and its capabilities are in relative infancy. The power of AI is already allowing for optimisations in the support space, and I think AI is really going to unlock a lot of untapped potential from a reporting standpoint. I look forward to how we can use this to continue to provide the best service experience possible for our customers.