5 on 5 would not chat again.
Thinking again about the Most Automatable Use Cases.
On yet another episode of checking yourself before wrecking yourself…
Trust is a foundational, if not THE foundational, rail of relationships.
Yet one of the first things that companies seem to be automating is the moment trust is earned/built/shattered.
Customer support jumps out as a cost center on the P&L. So, customer support is being firehosed with AI so ops managers and finance folks who are running for re-election are chomping at the bit to claim automation victories.
But customer support (when done well, or even mediocre, tbh) is often a lynchpin (and sometimes the only) human experience with a product. So, like it or not, customer service is a bastion of LTV, satisfaction, and retention.
Not to mention the (imo) underappreciated currency that comes by word-of-mouth when (at least these days) experiences barely clear the Acceptability Bar.
Get CS right and you keep your friend and meet theirs, too.
Get CS wrong and, well, you know.
Remember, stories are sticky.
The mechanisms by which trust is built (predictability, met expectations, effort absorption instead of offloading) are exactly the mechanisms AI-mediated support is least equipped to deliver.
Companies (especially those with declining users and/or engagement) know customer trust matters but they do the thing that erodes it anyway.
Classic short-term outcome prioritization over long-term gain? How dare I. Seems like it.
A surefire way to a trust liquidation sale.