What AI Adds to Customer Journeys
By Russell Bourne
Surely you’ve noticed the explosion of interest in 2023 about AI. Interest in it is everywhere: not only in the Customer Success industry, not only in tech, and not only in business.
Accompanying the massive interest is a mad scramble to figure out how to leverage it best. Some users are experimenting and failing fast - good, but sometimes at a price. Others may be hanging back to see how things unfold - safe, but slow.
Since this is a Customer Success blog, I’d like to share a few observations to help you decide if and how you should jump into the AI pool. As a disclaimer, I’m far from being an expert in AI but am fascinated at how we can use it for good across our Customer Journeys and Playbooks.
AI IS NOT JUST CHATGPT
First, let’s clarify that ChatGPT is only one small example of AI. I bring this up because I often see the terms used interchangeably. ChatGPT lies under a technology category known as the Large Language Model (LLM), which itself is only one example of several applicable AI types. The basic idea is LLMs can generate language-based content - a fancy term for text - intelligently.
A word to the wise here: if you’ve ever participated in your company’s leadership planning meetings around a new stack component or an upcoming new hire, you’ve probably heard some variation on “The problem we’re discussing will be solved when we get _______”. Honestly, that’s a bail-out on confronting a complicated problem. I’ve personally witnessed this said about Salesforce, billing systems, heads of Marketing, and so on. As you may have guessed, the arrival of these things didn’t, in fact, fix the problem. I bring this up because I’m hearing it about ChatGPT (or AI, interchangeably) now, and I caution everyone from falling into the trap. AI is not going to solve your problems just by existing. You have to use it in a specific, targeted way, and you must figure out what that is for your company’s use case.
WE’VE BEEN UNDER-USING AI FOR YEARS ALREADY
Generative AI is one example of a shiny new object in the AI world, but its use can be amplified if layered on top of already-existing, less-utilized methods.
For years, we’ve had simple speech-to-text technology - we’ve probably all had to remove notetakers from virtual meetings we didn’t want recorded. Many speech-to-text platforms come with out-of-the-box add-on functions that help users do things like identify key buying indicators, call out when important information is being shared, ensuring the right balance between talking and listening, and (related) sending users a warning that the other person on the call has developed a negative attitude or otherwise tuned out of the conversation.
These functions only scratch the surface of the insights GPT can give you. Imagine if your data pool consisted of transcripts of every sales call, every CSM call, and every Support ticket your company has ever had.
Internally, you could limit the GPT data source to that pool, and ask it for insights about key win/loss reasons, FAQ, and so on.
Externally, you can leverage Generative AI either reactively or proactively. For example, you could let customers self-serve without even knowing it by using it on the back end of a ticketing system or chatbot. Or you could use Generative AI to surface opportunity or risk, then use Natural Language Processing to prompt customers in a targeted, prescriptive way. A CSM would simply need to approve the note before it’s sent out.
These customer-facing outputs could be text, audio, video, wherever your customers want to be met. Combining text from an LLM with generative video is our near future. If you were at the 2023 ChurnZero BIG RYG conference, you probably saw a demonstration by Synthesia in the opening keynote. Synthesia generates AI video, performed by avatars of either fictional or real people. The video is shockingly realistic and will only get better in the years to come, but the catch is today you need to give the video avatar a script to read. How amazing to think of AI-generated text read by an AI-generated avatar.
SCALED CS ISN’T JUST AI; AND IT ISN’T JUST DIGITAL, EITHER
As a Customer Success community, we’ve been deliberate at finding the right terminology for many parts of our work. Whenever we talk about scaling CS - which almost always means how to handle small revenue customers at volume - we tend to lump the small customers into what used to be called a “tech touch” segment on a CSM-less customer journey. That name seems to have fallen out of favor because it implies small customers aren’t worthy of human contact.
Now I see the term “digital CS” but I don’t love that either - as a Segmentation and Customer Journey nerd, there isn’t necessarily such a thing as a segment that gets all-human or all-digital touch. The truth is all customer segments will be on journeys that are a mix of human and digital touchpoints. Even your biggest strategic customers will get something digital; and often - though not always - your smallest ones might get human help. Lately I’ve seen “Scaled CS” and that seems to be a good name, as long as we’re all together on the definition.
All of the above being said, I’d like to remind everyone that AI will not take over all the digital touchpoints in a journey. If nothing else, AI-generated content probably needs to flow through some digital delivery platform. For example, a proactive GPT email can be generated by ChatGPT, but it still sends through an email marketing platform, or a customer prompt will still flow through a chatbot.
Remember too that Scaled CS doesn’t have to be digital. It can be 1:many. It can mean in-person events, office hours, the list goes on. The touchpoints are human-delivered, but they’re efficient because they’re 1:many. For those who are horrified we’re losing human touch, these ideas are a scalable way of keeping it.
I’ll leave you with a final thought: I’m aware nothing in this article is particularly groundbreaking. I’m not an AI expert and I haven’t introduced a shiny new journey touchpoint. If anyone is reading this thinking it’s a firm grasp of the obvious, fair enough; my goal in writing this is to challenge you to think about how much of the obvious stuff you’re fully gone-live with today. CS is just like anything else; the fundamentals have to be strong. I’m positive the companies who are already following best practices on Customer Journeys are in the best position to adopt AI in an impactful way.
The Success League is a global Customer Success consulting firm that also offers a CS Leadership Certification training. Within that program are individual classes on Selecting Tools and Mapping Digital Journeys. Visit our website for more on these and our other offerings.
Russell Bourne - Russell is a Customer Success Leader, Coach, Writer, and Consultant. In a Customer Success career spanning well over a decade, his human-first approaches to leadership and program management have consistently delivered overachievement on expansion sales and revenue goals, alongside much friendship and laughter. Russell serves on the Board of Gain Grow Retain as co-lead for Content Creation. He is passionate about equipping individual contributors and business leaders alike to lean on their Success practices to grow their careers and help their companies thrive. He holds a BA from UCLA, and in his free time plays guitar semi-professionally.
