So how do you use AI today? Everyone uses it differently, and there are a ton of companies out there that will charge you a lot of money to use it the “best” way. The only thing anyone seems to agree upon is that it’s necessary and if you’re not using it in some sort of way you’re going to get left behind.
I can say with 1000% certainty, that when I started this portfolio website, AI was not used the same way it is now.
This technology is getting smarter and advancing faster than most people realize. There are those that are using it for life hacks, and efficiencies, and all of that, but I want to come at this from the approach of a CS leader.
How can I use this daily to make my work not only more efficient, but better. How can I get better customer outcomes using AI? Because the rest of it is meaningless if churn is high, NRR is low and your expansion rate is stagnant.
Two things I know right now.
- I would not have written playbooks had I known that AI could do it better, faster and more efficiently. While the playbooks are good, I’m fairly confident AI can give you a 50% leg up on any scenario before you even take the first step. So while I have plenty of playbooks written for CSMs, I would suggest you use AI to write them first before you consider the unique situation at hand. In addition you could probably just spend time prompting to get the best playbook before you even bother “writing it” from scratch.
- I run almost every customer scenario through AI at all times. Use the data available to you to have another set of eyes review it. One much smarter than you. Let it analyze what you are seeing and give you some suggestions. You can choose to ignore it, especially if its wrong. But I’ve found its right on these things more often than not.
You should be feeding your data into the AI at all times to understand the patterns and the efficiencies.
Take your support tickets, your account data, your metrics and run it through the different models. See what it says, then verify it and spot check it.
This is not to do a financial audit, but to give you an idea of what is actually happening.
This is not some new novel idea, but I think we spend too much time speaking about efficiency and less time on customer outcomes and making sure you’re using the data at hand to make decisions on what is best for the customer.
That is what Gong.io sold most of us years ago. We’ll analyze what the customer is saying and add probability to it.
Well this is how you should be running your team.
Example time.
I put in all my support tickets, 400+ tickets over a 6 month time period, the tags, the content of the tickets, the resolution times and I asked for patterns and suggestions on improvements.
In around 45 seconds it gave me a strategy I was able to spend an entire quarter working on.
It confirmed some of my own assumptions. It allowed me to bring data to leadership conversations, and provided a path forward with a detailed plan for the quarter. I of course adapted it, and made it fit more seamlessly, but why start from zero?
What I found most interesting is how I’m approaching my work now. Running most scenarios through an AI model before I complete them myself. Even if you ignore the model, I don’t see why you wouldn’t run it through first.
I pay for this myself, its $20 a month for Chat GPT Plus. To me that seems so beyond worth it. I’ve heard Anthropic’s Claude and that Perplexity are great as well. I think $60/month might be a bit overkill, but to me that seems like a small amount of money to fund on your own for the additional support.
Why not use the tools available to you to enhance the work you’re doing? It should be interesting to see where things evolve over time, but I know 30% of the busy work the team did previously is going to start going away a little bit each day, week, quarter and year until eventually, we’ll be doing something entirely different.
Robert Statsky 