If there’s a Golden Rule in customer service, it’s that people want to be treated like human beings—which implies that it takes another human being to respond to them with the right tone. The advent of artificial intelligence, or AI, in customer service, however, doesn’t break that rule the way you might assume.
AI can encompass a wide variety of technologies, but it tends to involve tools that can think the way people do, and automate tasks for scenarios that are more predictable in nature or based on contextual data. While AI is poised to have a huge impact on many other areas of a business, including sales and marketing, its use in customer service may represent a particularly good fit, especially for Canadian small and medium-size businesses that need help in that area.
That said, the journey to AI in customer service should be thought through now so that SMBs get the most out of what the technology has to offer and ensure it brings value to its customers right away.
1. Align AI with Customer Service Averages
Customer service, like marketing and sales, is among the most metrics-driven parts of an organization. Given the critical nature of the work involved and the investment necessary to support and troubleshoot customer issues, certain thresholds have to be met. Sometimes measuring individual agents or teams of employees can get bogged down with what we might call “human issues.” If someone was sick or on vacation, for instance, that has an impact on customer service.
AI not only removes some of those issues but can be applied across several other areas. Look at some of the most common metrics and discuss where AI could play a role.
- Average reply time: Because AI can integrate seamlessly not only with contact centres but social media accounts, customer service becomes more of an ongoing, near-immediate conversation than what SMBs may normally offer. Try calculating what a difference of even 15 or 30 minutes could make in terms of repeat sales or reduced churn.
- Average first reply time: There’s nothing worse than listening to hold music while you wait for the next available agent, or experiencing social media silence while the company’s employees are doing things other than monitoring their accounts. Even if AI isn’t capable of handling more complex customer service issues, explore whether it could at least bolster your reputation with customers by being quicker about acknowledging they’ve reached out and offering some kind of initial response on next steps.
- Average handle time: Given the volume of customer service issues in some organizations, there is often a pressure to resolve issues as quickly as humanly possible. AI offers huge potential here, because machines can retrieve troubleshooting information faster and more reliably, in some cases, than the human brain. Start planning what might be the most fundamental, frequently asked questions your customer service team needs to handle today, and where AI could shorten the cycle. Whereas average handle time refers to a single interaction, you’ll also want to look at average resolution time, the period it takes to completely close a case.
2. Optimize AI for Omni-Channel Journeys
Like their larger counterparts, SMBs have come to recognize that customers may interact with them through many different channels, from in person to online, via social media, or by smartphone. It’s not always easy for a smaller organization to be everywhere at once, but customer expectations are such that they aren’t always patient when problems with their purchases arise.
The best companies have already mapped out some of the most common journeys customers take, and ensure that if customers start an interaction in one channel, they can complete it via another. For example, someone may call a bank to deal with an issue but finish resolving it online.
Some questions you should start asking now include:
- Where is an in-person interaction with customers most critical, and where can AI fit in better? In some cases AI may be the better choice for customer service via social media, while for others a chatbot on their web site could free up team members to deal with in-person encounters.
- What kind of tone and depth is required based on the channel where AI might be used? For example, short responses could be automated for social media channels like Twitter, while a chatbot for a web site or smartphone app might need to offer more detailed suggestions.
- How should/could AI coordinate customers across channels? There is an art to shifting customers from a phone conversation to a web site where they can take advantage of self-service capabilities. Create a few examples of where AI might be able to predict certain kinds of troubleshooting scenarios and how customers could be more effectively routed.
3. Layer in Some Learning Goals for AI in Customer Service
Excellence in customer service requires training, not just an initial onboarding or orientation for new employees but ongoing efforts to ensure the team can keep up with new products and services, new features, or insights about common support issues. The challenge, of course, is that ongoing training for people often requires taking time out from actively dealing with customer questions or complaints. AI, on the other hand, can be “trained on the job” if organizations have a process in place. Some suggestions:
- Create a framework whereby information entered into CRM or marketing automation tools can feed customer support AI systems. That way, the data the system draws on to tackle issues is as real-time as possible.
- Where can AI upsell or cross-sell? As SMBs plan their go-to-market strategies for new offerings, look for avenues where promotions, discounts, or other incentives can be introduced as cases are resolved and closed.
- Establish a periodic AI audit. Just as customer service managers regularly listen in on live calls in contact centres, once AI tools are implemented, there should be a regular effort to work through a typical issue, with some thoughtful notes to help staff evaluate whether the technology is achieving the right kinds of responsiveness.
If you want even more in-depth information about what the future holds, download AI for CRM: A Field Guide to Everything You Need to Know, a Salesforce eBook.
