6 Realistic Applications For AI In Customer Success Teams

Article by

Ankit Vora

Journalist @ customerfacing.io



March 20, 2023

The buzz around ChatGPT & other AI tools lately got me wondering:

How relevant is AI to customer success? What are the use cases for it? Will teams who don't use AI start to get left behind?

I interviewed CSMs to learn more about some of the daily tasks/challenges they had, which might be solvable with AI. They included:

  • Extracting insights from customer calls
  • Predicting and reducing customer churn
  • Identifying upsell opportunities
  • ... and more

Let's dive into 6 potential areas where AI has an application in CS teams.

1. Predicting customer churn

AI can help CS teams predict which customers are likely to cancel their subscriptions.

I talked with a dozen CS leaders while researching, and almost everyone mentioned that they're using software tools in some form to predict customer churn.

Some of those, like CustomerSuccessBox, are using AI to calculate retention probabilities.

Success teams can integrate a tool like CustomerSuccessBox with their existing systems, after which it gathers customer data like demographics, billing, product usage, feedback and survey responses in one place.

Over time, a machine learning model can start to understand which traits & behaviors are more likely to lead to churn. That lets you calculate a retention probability for each account. And with that, you can help CSMs direct their efforts where the model deems a customer to be 'at risk'.

While not so simple to implement, the benefits of using AI in churn prediction are great. With enough (clean) data, accuracy of predictions will increase, and CSMs will spend less time trying to figure out who to spend time with. Additionally, with more data on what is causing churn, you can take steps to start preventing it too. For example, if a specific feature usage rate dropping will consistently lead to churn, you can plan initiatives to boost the adoption rate of that feature.

2. Analyze customer calls with AI

Spending countless hours extracting insights from customer calls? 

Maybe you want to:

  • Identify the most common problems your customers are bringing up during their conversations with your team
  • Track early indicators of churn like competitor mentions
  • Or track the frequency of specific keyword mentions over a specific time period
  • Search for calls mentioning a certain topic, so you can find example cases to learn from

Instead of spending hours trawling through recordings, tools like Gong can use AI to extract valuable insights from customer calls. 

Another tool example is SmartKarrot's SmartKonversations. It works in a similar way, offering keyword analysis and call sentiments data.

Suppose many of your customers used the term “poor user experience” during their calls with your team six months ago. So, over the following months, your product team focused on improving the overall user experience. 

Now, if you want to track the frequency of use of the keyword “poor user experience” in your customer calls, you don’t have to spend hours listening to customer calls and taking notes. Instead, AI platforms like Gong can track the frequency of keyword mentions to provide insights into how effective your product team’s efforts have been.

3. Identifying upsell opportunities

Expansion MRR is one of the top KPIs for CS teams, but figuring out which customers have expansion potential is tricky.

AI tools like Staircase can help CS teams uncover sentiment, health, and engagement by automatically analyzing all customer communications.

Using AI to do that can lead to higher accuracy of expansion potential vs. a more 'traditional' customer health score.

Additionally, customer success platforms like Custify, Vitally, Gainsight & others help CSMs gather customer data from different sources under one roof. They can identify meaningful patterns, such as an increase in product usage, to help CSMs figure out which customers are good prospects for expansions. 

In the near future, it’s highly likely that more of these tools will use AI for enhanced predictive capabilities.

Irina Vatafu, the Head of Customer Success at Custify stated:

There is huge potential to improve our tools by using AI to help CSMs make data-driven decisions. To prepare for this, we need to focus more on having clean data, understanding the behaviors of customers that show us they are healthy, having a clear idea of which actions lead to health and which lead to churn, and setting the stage for us to collect this data in one place.

AI tools will further help companies identify the factors that motivate different segments of customers to upsell based on their past behavior and individual preferences. Using this information, you can create effective offers that they’re likely to find appealing.

4. Assistance in content creation

AI can help customer success teams in various ways with content creation, including but not limited to:

  1. Write email responses to customers.
  2. Adjust tone & grammar (especially in non-native languages).
  3. Create educational content for customers.
  4. Update knowledge base articles.
  5. And more!

Some of them (like article writing) are fairly straightforward. There's lots of resources out there that help people use ChatGPT & other tools for writing.

Here's a less obvious application for customer-facing teams.

If you sell to a global audience, you're likely dealing with customers who speak different first languages. However, you may not have the budget to hire customer success and support representatives from every country you sell to. 

With AI writing assistants like Wordtune, CS representatives can understand the nuances of different languages and communicate effectively with customers regardless of the languages they speak.

At the same time, you can use these writing assistants to increase clarity and conciseness, respond to customers’ messages, generate content for different types of campaigns, and so much more.

However, it’s important to note that these tools should be treated as writing assistants. It’s a good practice to double-check the content you generate using these tools and add a level of human touch, as they can’t replace the emotional intelligence and empathy that human representatives can bring to the table.

5. Self-serve support with chatbots

So many customer service teams are flooded with an insanely high volume of inquiries that can easily be avoided by implementing a self-serve, AI-powered chatbot. Some of the most common questions they are often bombarded with include:

  1. How do I use Feature X?
  2. How do I cancel my subscription?
  3. Where can I find the user manual?
  4. How do I reset my password?

If your customer service representatives are drowned in a sea of such repetitive questions, it’s time to unburden them by implementing an AI-powered, self-serve chatbot. 

Doing this will reduce the volume of repetitive customer inquiries and empower your CS representatives to focus on more complex requests. Roi Kiouri, Customer Success Leader at Upstream reported a staggering 50% decrease in support tickets volume after implementing an AI-powered chatbot on their platform.

Another big benefit of implementing a chatbot is that – it’s available 24/7 and can handle multiple customer inquiries simultaneously.

And if you have a dedicated knowledgebase, an AI-powered chatbot can help customers find the information they need more efficiently by looking up relevant articles or resources based on their input – further improving the overall customer experience.

Slite, an internal knowledge management tool recently released an AI feature called Ask. Instead of reading/searching your knowledge base, you just ask it a question, and the AI gives the answer.

How long before we start seeing that in customer support, too? 🙂

Aside from this, there are several other ways an AI-powered chatbot can help CS teams like:

  • Collecting customer feedback. If you want to request customer feedback from a specific segment of customers, you can give them a nudge using your chatbot. 
  • Offering onboarding support to customers – depending on their behavior. More on this later.
  • Providing personalized recommendations.

6. CSM team coaching

CSMs need ongoing coaching and training to help them succeed in their jobs, which are constantly adapting to new market needs and trends. AI can play a big role here in helping CSMs and CS leaders stay on top of things.

Going back to some previously mentioned tools:

a. AI can provide recommended tasks to new CSMs

Based on the success & data of other accounts, AI can recommend certain tasks that an inexperienced CSM might otherwise miss. Here's an example. 'Leads created' feature adoption is too low? That's a nudge to figure out a way to increase it in order to reduce churn probability and increase LTV.

b. Call intelligence tools can assist in calls, give managers data to review, and find useful onboarding resources

Using data from a tool like Gong, a CS manager could review a full week's worth of calls from their new CSM. It can help to analyze the topics talked about, and give insights into what they could do more/less of.

Additionally, you can use such tools as a searchable library of customer calls, to quickly find interesting & valuable conversations that should be watched during a new CSM onboarding process.

Can AI replace human coaching?

Irina Vatafu from Custify mentioned:

I would never let AI completely replace coaching and brainstorming sessions with my team, but I would definitely use AI to improve the quality and efficiency of those conversations.

At the same time, it’s important to note that even though training CSMs is a huge priority for CS leaders, they have a jam-packed schedule and may not always be available for around-the-clock support and assistance. Irina further stated

AI can also solve a major challenge: availability.  We are often busy with so many tasks, and our day can be so fragmented that sometimes our availability just doesn't overlap. AI engines can provide around-the-clock support to relieve CSMs in difficult situations.

What success leaders think about AI in CS

I interviewed several customer success leaders at different SaaS companies to learn what CS leaders think about AI in customer success. Allow me to share their responses:

1. Irina Vatafu, Head of Customer Success at Custify

“AI will impact both high and low-touch models. Both come with their own challenges. The challenge for the high-touch approach is to create tailored experiences, understand customer needs as accurately as possible, and develop the most effective playbooks to avoid potential problems. This requires advanced data analysis skills and good intuition. The results may not always be perfect, considering that these requirements usually involve frequent context switches and reprioritization.  AI can play a big role here: It can automatically generate workflows and trigger alerts when there are red flags for potential problems, and it can help CSMs make data-driven decisions more efficiently.

For low-touch models, the challenge is the quality of self-service and the availability of support. AI will be a critical factor here as well. Self-service training materials, automated workflows for proactive outreach, round-the-clock support, improved SLAs, and even the creation of tailored experiences based on previous interactions - these are some of the first benefits that come to mind, but the list can be very long."

2. Rio Kiouri, Customer Success Leader at Upstream

"The future of customer success is bright. But, it's not all rosy. It's true, AI is not a panacea for every problem in the customer success space. It's still in its infancy, and there are many challenges that need to be addressed before it becomes a mainstream tool for CSMs. However, it's only a matter of time before AI becomes an indispensable tool for the CSM, just like CRM was in its early days. AI can process data for insights that CSMs then use to make decisions. This frees up time for CSMs to focus on high-value activities like coaching and relationship-building with customers

The best, most proactive teams are able to help customers before they need to ask for help. Predictive analytics will allow you to target your customers at scale and respond to their needs. When paired with superb engagement tools, AI will bring the speed, efficiency and intelligence of your Customer Success team into the future."

I’m excited to see how AI impacts customer success in the near future. It’s an evolving space that will impact both high-touch and low-touch models differently. And as Rio mentioned, it’s still in its infancy. However, it’s highly likely it’ll become an indispensable tool for the CSM, just like CRM was in its early days.