Klarna Just Posted $1 Billion in Revenue With Half the Staff. Here’s the AI Model Behind It.
In Q1 2026, Klarna reported $1 billion in revenue – up 44% year-over-year. Revenue per employee hit $1.4 million. That’s four times what it was in 2022.
They did this with 3,000 employees. In 2022, they had 5,500.
If you’re not paying attention to what Klarna just did, you’re going to be blindsided by a competitor who is.
What Klarna Did (Step by Step)
Starting in 2022, Klarna began systematically replacing high-volume, repetitive work with AI agents. They didn’t announce a flashy ‘digital transformation initiative.’ They just started rebuilding the stack.
Their primary AI agent now handles the work of 853 full-time employees. It answers customer inquiries, resolves payment disputes, handles merchant support – around the clock, in multiple languages, with zero hold times.
Simultaneously, Klarna cut its workforce from ~5,500 to ~3,000 while growing revenue from roughly $450M per quarter to $1B per quarter. Revenue doubled. Headcount halved.
The $60 million annual savings from AI customer service didn’t all go to the bottom line – much of it got reinvested into product and growth. That flywheel is now generating $68M in adjusted operating profit per quarter, up from $3M a year ago.
Why This Actually Worked
First, they rebuilt workflows around AI’s strengths rather than bolting AI onto human-designed processes. Most companies fail at AI because they try to automate a broken workflow. Klarna rebuilt the workflow.
Second, they kept humans for what matters most: high-judgment, high-empathy moments. When a customer has a complex fraud issue or a major dispute, a human engages. AI handles everything else.
Third, they treated AI adoption as a cultural identity – not just a cost-cutting exercise. Their CEO literally used an AI clone of himself to deliver the Q1 earnings summary. That’s a signal to the market and the organization: we are an AI-first company.
I’m Mike Partners. I founded AiExpert.org because I believe the strategies behind billion-dollar AI deployments should be accessible to every business owner. Here’s how to put this one into practice.
Your action step this week: Audit your customer interactions for the next 5 business days. Every time someone on your team answers a customer question, log the question type in a simple spreadsheet. By Friday, you will have a ranked list of your highest-volume inquiries – and that list is your AI automation roadmap. The top 3 question types are where you deploy first.
The SMB Playbook
- Audit your customer interactions for the week. What questions do you answer more than 10 times? Those are your first AI use cases.
- Deploy a customer service AI agent on your website, email, or messaging channels. Tools like Intercom Fin, Sierra, or a custom GPT-4o-based assistant can be live in days – not months.
- Track the ‘handles/human’ ratio weekly. Your goal is to get AI handling 60-80% of inbound volume within 90 days. Everything above that is found money.
The Klarna model is real. And it’s now available to any business owner willing to build it.
Frequently Asked Questions
How did Klarna achieve 1.4 million dollars in revenue per employee?
Klarna quadrupled its revenue-per-employee metric by simultaneously growing revenue from $450 million to $1 billion per quarter while reducing headcount from 5,500 to 3,000. The AI agent handling the work of 853 full-time employees was the primary driver – it absorbed customer service volume that would have required hundreds of additional hires, allowing revenue to scale without proportional staffing costs.
What does it mean to rebuild workflows around AI instead of bolting AI on?
Most companies fail at AI because they try to automate a broken human-designed process. Klarna took a different approach – they redesigned their customer service workflows from scratch with AI as the primary handler and humans as the escalation path. Mike Partners emphasizes this distinction because it explains why some companies see massive AI ROI while others see minimal improvement from the same technology.
How fast can a small business deploy an AI customer service agent?
With tools like Intercom Fin, Sierra, or a custom GPT-based assistant, a small business can have an AI customer service agent live in days, not months. The deployment speed depends on how well you define your top inquiry types upfront. If you spend one week logging your most common customer questions, you can have AI handling 40-60% of that volume within 30 days. Deployment frameworks and tool comparisons are available at AiExpert.org.
What is the handles per human ratio and why should businesses track it?
The handles-per-human ratio measures what percentage of customer interactions AI resolves without human involvement versus those requiring a human agent. Klarna’s target is having AI handle the vast majority of inbound volume. For small businesses, the goal is reaching 60-80% AI resolution within 90 days. Every percentage point above that threshold represents cost savings and freed-up human capacity for higher-value work.
Is the Klarna AI model only for large companies or can small businesses replicate it?
The Klarna model scales down directly. A business handling 200 customer inquiries per week can deploy AI for its top 10 most common question types and realistically automate 60-80% of volume. The $60 million Klarna saves annually translates proportionally – a company with $2 million in revenue might save $30,000-$60,000 annually while improving response times and customer satisfaction simultaneously.
