Below Water Level

Many teams seem to be getting side-tracked with fancy AI tools that will automatically synthesize customer feedback. There seem to be many incentives driving this behaviour:

• The lower barrier to entry for anything that looks like generalized feedback patterns. If it is so easy to put out something without the hard work, why not?
  
• Even if teams are willing to put in the work, the lack of skills to dissect customer feedback in a principled manner means that teams just keep tying themselves in knots. Who wants that?
  
• When there is pressure from the top to just do something about this AI thing, text summarization of tickets seems like an easy win for the quarter.
  
• It makes everyone look and feel really good because LLM-produced language is just so much engaging and articulate.

AI summarization tools make the whole feedback management process seem very efficient. Certainly, the one step in the practice of summarizing customer feedback becomes fast. From a system-wide efficiency and effectiveness angle, things actually become worse.

Customer feedback is gold but AI summarization tools add more distance and abstractions between teams and customers. The more abstractions, the more diluted or insight-free are the decisions, the more time-lag to actual value and more late you are in the market.

There is no alternative to the hard, down-in-the-weeds type of effort needed for engaging deeply with customer feedback.