Adding AI to your support team can be daunting. Our advice? You’ve got this

The thought of upleveling your customer service strategy with AI is incredibly exciting – but can also feel a little daunting.

We’ve started to roll out early access to Fin, our breakthrough AI bot powered by GPT-4, OpenAI’s most advanced model, and Intercom’s proprietary technology. More than 10,000 people have joined the waitlist since our announcement in March, and we’ve had some great learnings from the beta so far.

You might be asking yourself where to start with something that’s this new and (here comes that word again) unprecedented. Well, we believe customer service leaders already have all of the knowledge and expertise they need to implement AI and uplevel their team’s success.

“Even when it comes to AI, one of the most exciting topics in the world right now, it’s important to think about why and how you will use a new tool – and how it fits in with your team and strategy”

As a support leader, you’ve welcomed new team members lots of times. Much the same principles apply when adding an AI tool like Fin to your tech stack. We’ve laid out the most important things to consider when welcoming a new teammate – and how you can apply your experience to bringing a new AI bot on board.

1. What’s the job description?

When hiring a new support team member, you lay out their responsibilities in a job description, so that they – and the rest of the team – know exactly what they’re responsible for, and how they fit into the wider customer support strategy.

It’s a good idea to do something similar when you’re considering bringing an AI bot on board (although we wouldn’t recommend advertising on LinkedIn!). In our ultimate CS tech stack post, we wrote that a support team’s approach to their tech stack should be “strategy first, technology second.” Even when it comes to AI, one of the most exciting topics in the world right now, it’s important to think about why and how you will use a new tool – and how it fits in with your team and strategy.

Chat to your team about what you want to achieve by implementing Fin. What gaps could an AI bot fill? What new efficiencies could be unlocked with a tool as powerful as Fin?

Ask questions like:

• Where do we see an AI bot bringing most value to the team?
• How much time do we spend answering repetitive questions, and what would we rather spend that time on?
• How do we want to start incorporating an AI bot into our existing workflows? Are we ready to implement Fin across the board or should we try a more phased rollout?
• What kind of queries would Fin be best suited to answering, and which require a human touch?

“As with every new tool in your tech stack, monitoring, reflecting, and iterating on Fin’s responsibilities and performance will be essential”

2. Decide what success looks like

It’s good management to let a new teammate know what you expect them to achieve in their first week, month, or six months. Benchmarks and goals are vital to understand if a new starter is succeeding in their new role, or needs more support – and, when it comes to Fin, for you to monitor its performance and understand where you might need to focus more attention.

Decide what success looks like before bringing your new AI tool on board so you can measure the value it’s bringing to your team.

1. First, decide which metrics will be the most important for measuring Fin’s value. It could be automated resolutions, average response time, CSAT, deflection rate – whatever makes sense for your team.
2. Once you’ve selected the relevant metrics, make sure you’re aware of the team’s benchmark figures so you’re prepared to measure team performance with Fin against team performance without Fin.
3. As soon as Fin is set up (a super quick process), start comparing results!

3. Make sure the AI bot – and your team – are set up to achieve that success

It’s a manager’s job to ensure new teammates are given the resources and tools they need to succeed. Fin is incredibly easy and quick to set up – but there are things you can do to prepare yourself and your team for working with an AI bot; and to ensure Fin offers real value.

Fin draws all of its answers directly from your help center or a public URL. That means you can trust it to provide accurate information that’s limited to the content you choose – but it also means you need to be satisfied with the accuracy of the information you want Fin to share. Are the most commonly asked questions answered clearly and comprehensively in your help center or knowledge base? Take some time to:

• Fill gaps in your knowledge base to cover all the types of queries you’d like Fin to handle.
• Update and optimize existing content to ensure it’s clearly written and accurate.
• Create new content using commonly used saved replies (macros), and knowledge that may be widely known among your teammates, but not officially written down in a help article.

If this seems like a big job, don’t worry. Many of our customers are taking a phased approach to rolling out AI support within their support teams, pointing Fin at a small subset of questions or topics while they ensure their teams are ready for a full rollout. Check out this blog post for tips on optimizing your help content for Fin.

4. Give your team time and space to get to know Fin

Fin is a great conversationalist, but it only covers specific topics (your support content), and it can’t sit down for lunch with you and your team – not exactly a social butterfly. So, you need to find other ways to help your team become familiar with the new addition and get involved in optimizing for its success.

“[Your team] will likely find new, creative ways to implement Fin in their everyday work”

Rethinking team processes, responsibilities, and performance metrics to incorporate an AI bot will be an exciting opportunity, but also a challenge. Some of your teammates may be understandably concerned about adding an AI bot to your tech stack, and patience, understanding, and curiosity will be key as your team figures out the best way of working in this new world.

As your team gets used to AI support – spending less time on repetitive queries, and more on their favorite parts of the job – they’ll likely find new, creative ways to implement Fin in their everyday work. More importantly, they’ll establish clear areas where human skills and expertise are best suited to provide the quality of customer service they aspire to, such as:

• Complex, emotionally charged customer queries
• Bot orchestration, management, and performance analysis
• Support team operations
• Spotting gaps in your help center content that need to be filled
• Collaboration and feedback loops with other customer-facing teams

… and so much more. We’re excited to learn with you as your teams turn their hand to activities that develop their careers and move the needle for your company.

5. Check in often to make sure Fin is managed effectively

You’d never spend time onboarding a new teammate just to leave them to their own devices after a couple of weeks on the job. As with every new starter in your team, monitoring, reflecting, and building on Fin’s responsibilities and performance will be essential.

Consider not just the metrics you’re seeing, but also your team’s and your customers’ feedback on potential improvements that could make your team processes, help center content, or bot orchestration more effective.

Set time aside to:

• Check how Fin is performing against your benchmark metrics, and against you and your team’s expectations.
• Find opportunities to expand Fin’s remit by checking in on Fin’s conversations in Inbox. Are there questions Fin couldn’t answer that could now be automated? Create, update, or expand on your support articles as you go.
• Check your Custom Answers to find out how individual answers are performing, and update or tweak as needed.
• Collect your team’s feedback; discuss any ideas they might have for new and creative ways to employ Fin to take repetitive queries off their hands, or important areas they’d like to explore with the extra time Fin creates.

You’ve got all the tools you need to make the most of AI support

At Intercom, we believe AI support will make the jobs of customer service experts more varied, interesting, and fulfilling. And the best part is, you’ve already got all the skills you need to embrace what’s coming. Find out more about preparing your support team for AI at our upcoming webinar.