saved
I Haven't Lost a Customer Service Fight in Seven Months
Hraness cites a source capture. The source author remains the source.
gist
Dimitri Sudomoin turned a missing Annie's cheese packet into a seven-month, ten-for-ten consumer-claims pipeline with Claude that recovered over $12,000—from a $10 coupon to a $6,813 IRS penalty abatement. The agent absorbs the research and open-loop cognitive load, then runs multi-front campaigns (regulators, arbitration clauses, publicity) when frontline support stalls. He argues companies design for drop-off funnels that personal agents now break, even as support bots automate the other side of the arms race.
ideas
- Eat the cost until agents make pushback cheap. Small failures used to lose on time and cognitive load; a dossier-plus-branch plan flips that ROI.
- First-pass advice is weak; campaigns win. FTC form letters rarely move companies—arbitration fee leverage, home-state regulators, and public reviews do.
- Context, not cleverness, is the agent's job. Voice dumps, OCR, warranty research, and live call coaching clear the open loop so denials become the next planned branch.
- Both sides now run agents. Support bots apply stall scripts more consistently, so vanilla ChatGPT tips deflect and precision filings matter more.
- Spare small businesses; target the funnel. Point the same machine at reviews for good local work, and reserve escalation for corporations that starve support.
quotes
“Most of what the agent does for me is manage the context.”
“What actually wins is a multi-front campaign.”
“An agent breaks the funnel.”
“AI agents have infinite patience. They will fight for your claim for as long as you keep paying for your Claude/ChatGPT subscription.”