Klarna Replaces 700 Roles with AI. How Should Leaders Approach AI Workforce Planning?

Which part of your company will still exist five years from now? How do we execute workforce planning when AI may severely disrupt us?

I’m hearing these questions from leaders across industries and countries as the news hit that Klarna, a Swedish fintech company, replaced 700 customer service jobs with AI.

Let’s dive in.

Klara lets AI do the work of 700 customer service jobs  

Klarna is a ‘Buy Now Pay Later Solution’ that services 150 million global active users and 2 million daily transactions by partnering with retailers like H&M, Saks, Sephora, Macy’s, Ikea, Expedia Group, Nike, and Airbnb.

On February 27, Klarna issued a press release that went viral on social media. 

In the release, it shared how it launched an AI customer service agent powered by OpenAI that, in its first month, had 2.3 million conversations, two-thirds of Klarna’s service chats.

While impressive, the second bullet point in the statement had people going wild: “The AI is doing the equivalent work of 700 full-time agents.” 

Expanding on this, Klarna shared that the quality of the conversations is on par with humans and that the AI did a better job in issue resolution, further reducing the number of conversations needed. 

The AI also impacts time to resolution: people contacting Klarna get their issues resolved in 2 minutes versus 11 minutes with human customer service. 

All of this would lead, as Klarna shared, to a 40 million dollar profit improvement in 2024. And with a $2-3 million implementation cost, that’s a pretty dandy ROI. 

‍

SPONSORED BY:

Retain Your Top Talent [New Insights + CxO Roundtable]

Even with AI replacing many jobs, you’ll need your most senior talent to drive the business.

Join C-level executives from Ogilvy, Standard Charted, Culture Amp, and a range of companies across industries to hear new insights on how to retain your top talent.

Zoom online event, March 22, 9AM Singapore Time. 

Request your ticket here.

(CxO’s only. Ticket request does not guarantee entry.)

Not so fast, Klarna

Fairly quickly after the news went viral, commentators placed some critical remarks.

First, people highlighted how Klarna raised $800 million in funding at a valuation of $6.7 billion in 2022, only to see that valuation drop by 85% after losing $580 million in the first half of 2022.

These losses moved Klarna to lay off 100 employees in September of that year and to place a bigger bet on AI servicing their 150 million users. While Klarna denied that there is a causal relation, that seems iffy given that they mentioned AI when executing these layoffs.

People also shared that the timing of this announcement, especially the $40 million profit improvement, is curious as Klarna is rumored to go public this year and needs to raise its potential. 

Klarna itself backtracked a bit on the original statement, saying, "we chose to share the figure of 700 to indicate the more long-term consequences of AI technology, where we believe it is important to be transparent to create an understanding in society.”

In short, AI may be doing the work of 700 people, but Klarna had no intention of hiring them. These are “AI jobs,” which only partially impact people.

Customer Service and AI

Regardless, people will be impacted. 

As I mentioned in my article covering the 6 AI Trends for 2024, Forrester predicted that customer experience would improve significantly thanks to AI. 

Generative AI would augment customer service agents to “improve top CX drivers, such as answering questions faster and better, resolving problems on first contact, communicating clearly, and leaving the customer feeling respected.”

This builds one of the earliest studies on Generative AI and customer service by Stanford research associate Erik Brynjolfsson early last year.

In that study, AI delivered value by analyzing call center workers’ calls with the highest satisfaction score and working backward to upvote the best answers in real-time, allowing even lower-skilled employees to boost the quality of their answers by up to 35%. 

But beyond improving the experience, if the Klarna numbers are to be believed, AI can handle more conversations and close them quicker without affecting their quality. 

Therefore, we can expect reductions in customer service teams, perhaps more than any other category of jobs AI will replace. 

And the market is noticing. Klarna said to work with “four to five large third parties that collectively have over 650,000 employees.” 

Well, one of the biggest customer support providers, Teleperformance, saw their shares fall by 29% after the Klarna AI news broke, with no sign of recovery.

Especially because this is such a real, tangible example of AI potentially taking over jobs, as Klarna’s CEO said to CNBC:

“As much as we see about AI and high productivity, I seldom feel like I see something that has had this practical implication. it is not only a prototype and a demo, but it is live, covering 23 markets and 35 languages. Obviously, it will have a massive impact.” – Klarna CEO Sebastian Siemiatkowski.

Integrating AI in Workforce Planning 

There’s no way to avoid it.

We now have three major studies, from McKinsey, the IMF, and Goldman Sachs, all pointing towards AI replacing around 40% of jobs. And all of them have Customer Service as one of the key areas where AI disruption will be swift, as this McKinsey chart shows:

Additionally, research from Bain I’ve quoted before shows that productivity goes up by over 41%.

In short, the helpful case study from Klarna shows the writing on the wall. While their team tried to downplay the effect of AI on existing jobs, in a previous interview with Fortune, Klarna admitted it did specifically look at AI to reduce headcount costs:

“We are in the fortunate position of being a growing company so for Klarna, AI enables us to grow more quickly without adding headcount as quickly as we would have done previously,” – Klarna

Now with a clear case study on how to let AI do human jobs, more leaders will wake up to the ‘opportunity,’ shared Helena Turpin, Co-Founder of GoFIGR, an SME talent opportunity and insights platform, for this newsletter.

“Even though the Klara AI may not be that impressive yet, many CEOs may rethink their workforce planning following Klara's viral announcement." – Helena Turpin, Co-Founder, GoFIGR

And indeed, new Slack research shows that for 81% of executives say building AI into their organization’s workflows is an urgent priority, and around half say it’s highly urgent. 

Organizations can’t afford to fall behind, added Helena.

"My hypothesis is that for many leaders, the impact of AI is too big, vague and hard to incorporate into workforce planning. But it worries me that it’s a blind spot; businesses are going to miss out on opportunities and risks."

Suzy Chief People Officer Anthony Onesto agrees and highlights that workforce planning for the age of AI is a ‘new normal’:

“This is the new normal. Markets reacting to cuts (ala Meta), revenue growth inspite of those cuts, now layer on AI and automation. I think the story is that we are now in perpetual reductions in force that historically would impact revenue growth, but that is being supercharged by AI. So investors, boards and CEOs are waking up to the reality that they can cut labor and grow. The holy grail of business.” – Anthony Onesto, Chief People Officer, Suzy

But, while urgency and intent are high to keep up with competitors including future “AI unicorns of one,” practically, it seems like many companies are falling behind. As the Slack research shows, 43% of workers say “they lack any guidance from leadership about how to start using AI in their jobs.”

This is in line with what Rebecca Hinds, Asana told me in our interview: there are huge gaps between executive and employee experiences on AI adoption:

“The second gap we see we call a transparency gap. When we ask executives whether they've been transparent in terms of how they're using AI and how they're bringing AI into their organization, executives are much more likely to say that they've been transparent compared to what individual contributors say.”

These numbers show that there’s an imperative for companies to not only form a AI workforce strategy, but to also take their employees on that journey and transparently build a plan for the future together. 

Planning for a human + AI workforce starts by thinking about the tasks to be done, rather than the roles required, as Thinker 50 honoree Dave Ulrich shared in our interview. Break down all that a company has to accomplish to the task level, and sort them by what human should do, and what AI can take over:

“People like to talk about workforce management. What are the people that have to do it? Are they full-time? Are they part-time? Are they fractionally remote? I think we're going to see an addition to that full-time, part-time contract. A lot of the work can be done through AI bots. And so what do you then start with to say, what are the requirements of this job? And here they are: do I need a full-time job for part-time work in technology? Do it, and a lot of the work in our field, I think, will be done through technology.” – Leadership guru Dave Ulrich

The Bottom Line

We’ve decidedly moved from AI disruption as a research topic to it happening right in front of us, in the real world.

The Klarna case is just one example of companies being able to replace humans with AI. 

Tyler Perry shut down an $800 million studio expansion after seeing Sora, saying “I don’t know how we’d survive.” Apple TV director James Hawes advised the UK parliament that he envisions AI taking over all facets of TV production, including actors. 

In architecture, according to Goldman Sachs one of the industries most likely to be disrupted by AI, work that would take weeks now takes second. 

In a recent webinar, Archetype’s Kiem Nguyen showed how tools like Prome and Maket that can do work like creating detailed visualizations, full-blown renderings, and soon even technical drawings, in minutes rather than weeks. 

In short, AI is impacting all of us, across industries. Our AI Top 100 shows over 60 platforms with more than a million monthly users, a large share of which are used for work, from coding faster to training sales people in real-time.

PR talk about ‘it’s about people + AI’ starts to sound hollow in a world where it’s become too obvious that AI can take over our work without quality loss. As AI thinker Dave Shapiro shared, you can’t lay of 42,000 people when you’re highly profitable and then say it has nothing to do with AI.

In the longer run, these layoffs are not a bad thing, at all. Leaner organizations run better and are typically more enjoyable to work for, as HR Tech guru Josh Bersin highlighted:

“Over-hiring results in bureaucracy, slowdown, and loss of innovation. Just as Meta has learned this, I believe every company has to learn this lesson. In today’s talent-constrained global economy, it’s time to run our companies with a focus on “fewer people” and “more clarity.” – Josh Bersin

Recent PwC  research showed that CEOs believe 40% of time at work is wasted, with email, meetings and recruiting (hello, AI Recruiting!) as some of the biggest time hogs.

Repack your organization with the core talent you need and leave the rest to bots. That makes work for the rest of us a lot more human!

As one Redditor commented on the Klarna story, “It used to be that they had to pay an entire human to fail to serve their customers. Now with AI they can fail to serve you for at a fraction of the cost.”

– Daan

See more:

• AI Statistics
• AI in the Workplace
• Jobs AI will replace
• How to use AI (for work)
• Top AI tools
• AI in HR
• AI Recruiting
• AI Recruiting Software
• AI in Management‍
• AI Websites
• ‍Free AI Tools
• ‍AI Productivity Tools
• ‍AI Presentation Generators
• ‍AI Accounting
• ‍GPT's Prompt Generator