"AI is helping us become leaner, faster and more responsive to what our customers care about, leading to a much, much better experience," says David Sandström, CMO at Klarna. "And we’re actually driving more marketing activity while saving tens of millions of dollars a year."
In May 2024, Klarna announced how they were implementing AI and how they were able to reduce marketing agency costs while running more campaigns. Recently, AdWeek covered a piece stating Klarna launched 30 campaigns with 100% AI Generated creatives, which performed so well that it helped Klarna cut their annual marketing budget by 12%.
But let’s take a step back and look at how things fell in place resulting in today’s buzz.
The Massive Scale of Klarna
Started in 2005 by Sebastian Siemiatkowski, Victor Jacobsson & Niklas Adalberth, Klarna (Kreditor then) was one of the pioneers of the BNPL model – Buy Now Pay Later based out of Stockholm, Sweden. It brought the BNPL model to early adopters of online shopping which quickly picked up the pace and was expanded to 18 markets globally. By the end of 2023, they had 3800 full time employees with $2.5 Billion in revenue.
Now, Klarna serves 85,000,000 active users, used by 575,000 merchants to receive payments, facilitating 2,500,000 transactions per day.
Let’s dive further into Klarna’s situation before and after integrating AI
Klarna spent SEK 210.5M ~ $191M on their sales & marketing last year.
One thing stands out from this P&L report - there is a stark difference of SEK 33.7M ~ $30.5M between the two - Q3' 23 and Q3’24.
Apart from this, Net income reached SEK 216 million in Q3 2024, a 57% improvement as Klarna continues to compound growth and enhance efficiencies through AI innovation. Net income (loss) for the first nine months of 2024 was SEK (116) million, a 94% improvement compared to the same period last year. Revenue increased by 23% in the first nine months, driven by a 16% growth in Gross merchandise volume (GMV) and an increased revenue take-rate.
But how did AI actually help Klarna in saving $10M out of this $30.5M from their marketing expenses?
Adopting AI Early and Turnaround
Early Adoption
When ChatGPT launched in 2022, Sebastian saw the potential early on and decided to start incorporating AI in their company and signed license commercials with OpenAI. Klarna became one of the early adopters of ChatGPT. And since then, they started experimenting with the application of AI in their departments.
And since then, employees of Klarna have built over 300 custom GPTs for internal usage specific to their manual tasks that had potential to be automated.
Let’s look at some use cases of AI at Klarna’s –
Use Cases
- AI handles 2.3M chats - 67% of Klarna's support volume.
- Works like 700 full-time agents.
- Customer satisfaction matches human agents.
- 25% fewer repeat questions than humans.
- Problems solved in 2 mins vs previous 11 mins.
- Serves 23 markets 24/7 in 35+ languages.
Klarna published these results in February 2024.
- AI has helped Marketing team save $10 Million on an annual period where $4 Million were saved on the costs of marketing agencies whereas ~ $6 Million on content production with the help of GenAI models like MidJourney and Dall-E.
- AI made 1,000+ images in 3 months, cutting creation time from 6 weeks to 1 week, with all quality checks included.
- AI writing tool handles 80% of Klarna's marketing copy.
- Image creation time: 6 weeks → 1 week.
Just 2 months later, Klarna launched these results in late-May 2024.
AI-powered shopping feed is developed on the large databases of price, products & services of Klarna and serves through 5 functions:
- Search: Chat to find products - like sulfate-free shampoo. See reviews, prices, and buy directly.
- Recommend: Get personal suggestions based on your shopping habits and brand preferences.
- Compare: Find best items in your budget with AI-created pros and cons list.
- Research: Ask anything about products - from ski gear to rollerblading equipment. AI analyzes and suggests.
- Find deals: Access prices for 5.6M products across stores, plus stock and delivery details.
Klarna launched this AI-Shopping assistant in mid-September 2024.
It’s evident that Klarna has been consistently testing and deploying AI use cases in the form of agents and assistants in every single department across the org right since January 2024. And it shows!
Saving $30M a year and beyond!
This is just the start. There is a ton of marketing and user data available where AI can be leveraged to take data-driven decisions and build intelligent automation which is relatively a more difficult problem to solve and probably, even higher-impact.
Some of the decisions that can be taken faster with AI and automation:
- How to scale the marketing efforts efficiently?
- How to reallocate existing budgets across platforms and campaigns?
- Which creatives, positioning, and messaging works for each segment?
- How to engage with new audiences?
The fastest-growing marketing teams are adopting AI
Just like Klarna, several fast-growing brands and agencies are experimenting and adopting AI to improve efficiency, customer experience, and scale faster than ever.
Here are some of the high-value use cases that we’re observing:
- Reporting & Performance Summary
Most marketing teams spend 20% of their week (~5-10 hours per person per week) on creating reports for the CMO, CFO and other management. This could be 90% automated and potentially save thousands of dollars in repetitive tasks. Instead, this time could be spent on improving performance & optimisations
- Creative Analysis & Recommendations
Creatives have turned out to be the biggest (and perhaps the only) lever in performance marketing. Most ad platforms are moving towards a setup which only needs you to experiment with your creatives. To align with this trend, you need to create and test hundreds of creatives. The only way to do that currently is to manually scan each creative and guess what “could” be working.
With AI, you can find granular patterns in creatives to replicate what works and refresh creatives before fatigue sets in.
- Budget Allocation & Scaling Strategy
The most pressing question for CMOs is figuring out how to allocate the marketing budget across platforms, audience segments, product categories and funnel stages. The impact of this decision definitely lasts for more than a quarter and it's absolutely crucial to get this right. The beauty of LLMs is that they can create scenarios not just based on historical data, but based on real-world events (like the US elections or Black Friday) and business context (like KPIs, constraints, targets, etc.)
If you want to do similar things within your marketing team, schedule a 30 min free Ad Account Audit + Consultation with us.
