Don't make the
same mistake
as Klarna.
Klarna cut 700 customer service roles between 2022 and 2024. By mid-2025, they were rehiring. Customer satisfaction had collapsed. Most boards are still pushing for AI to reduce headcount — evidence now shows this is the wrong target. Sustainable savings come from designing the human-AI combination correctly from the start.
of companies that replaced staff with AI will rehire by 2027.
Source · Gartner, 2025Five things
in this recording.
-
01Why headcount reduction is the wrong unit of measurement for AI ROI
What boards are asking for isn't what delivers sustainable value. We show you what to use instead, and how to make the case for it.
-
02What the Klarna case and Gartner's 2027 prediction tell us
Where AI falls short when it replaces humans rather than augmenting them, and the pattern that almost every organisation repeating this mistake shares.
-
03Where AI genuinely drives cost reduction in CX
The specific workflow and volume types where AI delivers compounding savings without sacrificing satisfaction, from real enterprise deployments.
-
04How to structure people and AI together so savings compound
The design principles behind human-AI combinations that work, and the sequencing that separates sustainable savings from expensive mistakes.
-
05What to tell your leadership when they ask for cuts your AI strategy can't support
A practical framework for the board conversation — how to reframe the target and build a savings case that holds up to scrutiny.
One number.
Tells the whole story.
By mid-2025 they were rehiring. Customer satisfaction had collapsed. The savings weren't sustainable because the human-AI combination wasn't designed correctly from the start.
Source: Klarna, public announcements 2022–2025The companies finding sustainable savings are the ones designing the human-AI combination correctly from the start — not the ones reacting to board pressure with knee-jerk headcount cuts.
Three ways
boards get burned.
CSAT collapse
Cutting human roles before AI is ready to absorb them leads to customer satisfaction falling — exactly what Klarna discovered when it scaled back its AI-only approach.
Expensive rehiring
Rehiring costs — recruitment, onboarding, retraining — often exceed the original savings, with a 12–18 month lag before performance recovers to baseline.
Board credibility lost
A failed headcount-first AI strategy damages internal confidence in AI programmes — making it harder to secure investment for the right approach next time.
Two executives
who've seen it firsthand.
.svg)
Roy has worked with enterprise contact centres deploying voice AI at scale and has seen first-hand what happens to customer satisfaction when the human layer is removed too early.
Kane has spent a decade advising enterprise organisations on conversational AI strategy and has interviewed hundreds of practitioners on what actually works in production.