A slick AI demo can make almost any customer service problem look solvable.
Unfortunately, it can also be where the trouble starts.
CX teams are under pressure to automate faster, reduce demand, and show a return on their AI spend.
Under this strain, many are simply implementing AI for the sake of implementing AI. And even those who try to be more strategic in their approaches often fall into the trap of simply identifying a busy support channel, introducing a chatbot or agentic tool, and hoping the numbers move in the right direction.
But a higher containment rate does not necessarily mean the customer got what they needed.
The bot may have answered quickly, failed to resolve the issue, and sent the customer back through another channel. The interaction disappears from one dashboard and reappears somewhere else.
Craig Crisler, CEO and Co-Founder of SupportNinja, believes too many organizations are putting the technology before the customer problem:
"You start with the problem or the strategic goal you’re trying to achieve, then you find a solution, and then test against that to understand whether or not it works.”
That is a considerable shift from the way many AI customer service projects have been approached. Rather than asking where a chatbot can be deployed, CX leaders should ask why customers are getting in touch in the first place.
Finding the Friction Before Picking the Tool
When it comes to identifying issues, the answer is often sitting in data that businesses already hold.
CRM records, emails, interaction transcripts, customer surveys, escalations, repeat contacts, and sentiment trends can all expose points of friction across the customer journey.
AI can be useful here, too. Before it is asked to automate a process, it can help analyze large volumes of customer data and identify where an experience is going wrong, as Crisler explains:
“The beauty of AI is that you can analyze data at scale. Using that tooling to analyze first and understand, 'okay, well, this is a friction point in our CX journey. Let’s look at that and see if there’s something we can fix around that.'”
For example, if customers keep contacting a business to ask for an update on a claim, handling those calls more efficiently may improve operational metrics. It does not remove the effort that caused the calls.
SupportNinja has previously pointed to a benefits administrator that addressed this issue with proactive alerts across four stages of the claims journey.

