When I first encountered SaaS back in 2006, I immediately saw that one of the two sea-changes it offered was the ability for a software vendor to monitor what their customers were doing with their applications’ feature sets in real time. This was and is a gold mine of vital data in the continuing conversation between customer and vendor about needs and value that leads to retention. But I was shocked to discover some time later that vendors were not making use of this strategic resource, and most still aren’t today. Worse, in all the hoopla about the wonders of AI, I see a similar scenario unfolding. In the rush to add the all-important AI tag to applications, vendors are risking a strategic error in how they perceive the value of the technology. They’re leaving pure AI gold lying untapped.
On the surface, it seems almost too simple. An AI agent is created to engage in automated conversations with customers so that more expensive human resources can be applied elsewhere. Zap! Questions get answered in seconds. Of course, vendors need to periodically review the answers to ensure that the AI isn’t hallucinating, but after awhile, a reliable database of answers is verified. Automation can proceed efficiently and effectively. You can even analyze the patterns of those exchanges to reveal problems with the applications’ user interfaces or other errors. Customer has a problem, customer asks a question, gets the answer and, presumably, is able to implement that answer to solve their own problem. (Note the risks in that last assumption. What if they *aren’t able to implement the solution?)
Gold Mining
But the potential of AI is so much greater than faster answers to questions. By focusing predominantly on the speed and accuracy of the answers being provided, a greater wealth of knowledge risks going untapped. For example, what does it mean to the chances for long term retention and expansion if several important users from a SMB customer are asking a particular question 4 weeks after go-live? If you don’t have the data to make that analysis, what’s the cost?
OpenAI doesn’t maintain a database of the conversation transcripts itself; the application vendors have to do that separately themselves. As with the SaaS ability to see what the customers are doing with the application feature sets, there’s a gold mine of insight that AI can produce about your customers — but only if you design to take advantage of it. What other golden nuggets are being overlooked? If a customer asks a particular question, will your AI agent know to probe deeper to uncover unmet needs and expectations?
Taking the Lead
Customer Success, this is your cue to step forward to take the lead in visioning what your total product could become, drawing on your domain expertise and your in-depth knowledge of your customers. If you don’t know how to frame the discussion, let’s talk.
