AI Customer Service: Customer Support in 2021

By Ruth Favela

Customer service has long played a pivotal role for companies. Yet, having a well-organized customer support service or system is more crucial than ever in the digital age.

Why? Consider the modern consumer. They:

It may sound like a tall order, but it doesn’t have to be.

Artificial intelligence (AI) can transform your customer service by solving common questions, assisting with complex tickets, routing tickets, and searching for data, and more. If you intend to address the needs of the digital consumer in 2021 successfully, you need to consider AI-powered customer service.

Keep reading for a breakdown of pros, cons, additional ways AI improves efficiencies, and which customer support platforms utilize AI.

The Pros and Cons of AI-Customer Service

As noted, AI customer service is a powerful avenue for helping customers. However, there are two sides to every story. Here’s what we mean:

The Pros of AI 

  • Always on call. Time zones do not restrict automation! With self-service tools like knowledge bases, helpdesk, FAQs, and chatbots, your customers can get answers 24/7.
  • Little wait time. Speed is critical when it comes to customers. AI can help instantly by deflecting simple tickets, answering straightforward questions, or letting customers resolve issues themselves.
  • Access all your data. Most organizations have data stored in disparate sources. AI can surface information from your CRM to your ticket management platform. No more digging!
  • Reduce costs. Machine learning and advanced speech recognition improve accuracy, getting the correct answers to customers in need fast. Consequently, you save money. For every hour an AI can take off average handling times, a customer support team can save as much as $1 million annually.
  • Less human errors. Automation can offset a ton of manual data entry, which can decrease the risk of mistakes. It can also help categorize and escalate tickets to an agent with the proper knowledge or skill set to handle a complicated issue.
  • Improve efficiency. It’s important to note that AI will never replace human agents. It will only help them. AI can aid with triaging, answering Tier 1 tickets, onboarding, and training. 

The Cons

  • Setup. Implementation is often a scary word because it conjures hours spent learning a new system and determining processes. A lot of AI will take time and effort, but AI like Agatha can get you up to speed in just a few weeks. 
  • Accuracy. Effectively routing and triaging tickets requires deep knowledge of situational and historical precedence. Some systems tag tickets differently, and occasionally, not very well. Agatha, however, reports up to 98% accuracy in answering and routing tickets.
  • Up-keep. You will have to tend to your knowledge base and other self-service options to guarantee your information aligns with your current policies. However, with Agatha, you don’t need significant overhauls or a ton of upkeep to keep her relevant because she’s built on your data and can retrain when necessary. 

Depending on which customer service platform and AI solution you choose, the pros will outweigh the cons. So, what do you pick?

Customer Support Platforms That Use AI

Organizations today have access to an abundance of different customer service offerings, like Zendesk, Salesforce, and more IBM. All of these use AI to power their functionality. For example, IBM Watson Assistant is an AI-powered chatbot, which is slightly different from the run-of-the-mill chatbot you’ve likely seen on a website.

Actually, the word “chatbot” doesn’t do justice to what’s on the market today.

Take Agatha, for example. Agatha reads questions comprehensively, looking for homonyms, context clues, and other data that will help her answer questions more carefully and correctly. She uses Natural Language Understanding to match keywords within text to gain accuracy and create effectiveness. Additionally, she learns over time and won’t become obsolete

A more traditional chatbot that doesn’t rely on advanced AI can undoubtedly answer questions but not at the level as Agatha or others. She and other similar solutions will need a typical volume of 2,000 tickets a month to work wonders, but if you’re seriously contemplating these options, you’re likely at that rate, anyway.

With that in mind, here are two of the top-rated customer service software solutions available that leverage powerful AI reminiscent of Agatha: 

  1. Zendesk: Zendesk offers support in ways most convenient to your customers, including messaging (apps, live chat, social, email, and voice), help centers, community forums, and AI-powered bots. Zendesk is also known for its marketplace of integrations and plugins – including Agatha!
  2. Salesforce: Salesforce provides a robust customer service software called Service 360, powered by their AI known as Einstein. Like Zendesk, support teams can engage with customers on various channels, automate workflows, provide self-service, and utilize AI-powered chatbots. Salesforce also integrates with Agatha, extending your support capabilities.

Zendesk and Salesforce, while perhaps the best-known, aren’t the only available systems. They are great fits for Agatha, and a combination of these resources can amplify your efforts.

Additional Ways Improve AI Customer Service Improves Efficiency

By now, we’ve made quite the case for AI customer service. Let’s take it a step further and examine the specific ways it can improve you and your team’s efficiency:

  • Minimize wait times. Your customers don’t want to wait for answers. If you feel limited in how you can help your customers, that’s a clue to turn to AI. AI can automate email responses with detailed answers, resolve commonly asked questions, help your agents with macros and templates, or serve up an accurate knowledge base article. All of this can be done in just a few minutes, too.
  • Auto-triage. A full inbox or long queue of tickets is familiar to most agents. Managing these tickets can take a good chunk out of their day, preventing them from other work. Auto-triage functionality like Agatha’s can read and route tickets, remove spam, determine escalations, and send tickets to suitable agents.
  • Help more customers. In just a week, AI like Agatha can reduce your agent workload by up to 20%! With recurrent and Tier 1 tickets out of the way, you and your agents can focus on helping more customers that require a human touch.
  • Uncover trends. By gathering metrics on your tickets based on a detailed understanding AI provides, you’re learning about your customer support trends. You can see common questions and concerns to prepare your team. You can make informed decisions about your customer and product roadmap. You can even see individual agent performance to see how AI is helping, who’s doing well with it, and who may benefit from additional training.

Perhaps one of the best things about all of these improvements — at least, the one your CEO or operations manager will love? AI customer service saves costs, especially in these two areas: 

  1. Time. Without AI, agents have to sift through disparate information sources to find answers, from the simplest to the most complicated. With AI, particularly Assist, you can consolidate your data sources into one. That way, your agents can quickly locate the data they need, whether it’s in a help desk, wiki, macro, template, or somewhere else.
  2. Training. Customer support AI can enhance the onboarding process for new agents. AI like Solve can free up seasoned agents’ time to help train new agents. Assist can help new agents gain confidence by eliminating the time they spend doing manual research, which is especially helpful if they’re working on a complex issue.

Before taking the plunge, you need to know how to measure your investment. 

Measuring the Effectiveness of Your AI-Powered Customer Support

When it comes to the customer support and service industry, specific data points and they’ll be vital to you and your team as you navigate the AI waters. Try looking at:

  • Customer Satisfaction Score (CSAT) – a commonly used KPI to see how satisfied a customer is with your product or service.
  • First Contact / Auto-Resolution Rate  – the percentage of tickets or issues that are automated by AI.
  • Transactional Surveys – feedback after a specific interaction. 
  • Wait Time – how long it took for either your AI or an agent to respond to a customer.
  • Average Resolution Time – the median time it takes AI or an agent to resolve a ticket from the time the user sends it.
  • Volume by Channel – where your customers are reaching out. 

All of these are a blend of operational and organizational metrics. Both kinds are necessary to measure because they can guide your team to achieve maximum potential using your customer support platform and AI.

Wrapping it Up

You can see why having a well-oiled, modern customer support system in places matters in 2021. Still, there’s a lot to know when it comes to AI customer support. You have to take into account the pros and the cons. You need to research your solution options. And, of course, you need to know how to measure it all. 

Hopefully, we offered some answers to your questions and proved that a customer support platform backed by AI will save you time, money, and energy. If you need just a bit more information, please refer to our “Why Agatha!”


Forethought's guide for navigating the evolving landscape of customer support for 2021 and beyond.

Get your copy
Call to action decor Call to action decor