Klarna uses AI to make budget cuts

Klarna AI Agency Makes Budget Cuts

Jason Tabeling
Head of Solutions
Jun 5, 2024

Klarna has been raving about the productivity increases they have achieved with AI over the past few months.

It started in February when they announced that AI was doing the work of 750 customer service agents—the equivalent of two-thirds—within the first month. This work covered 2.3 million conversations and drove a $40M profit in the US. Pretty impressive… but it didn’t stop there. 

In May, they followed this up by announcing that they had cut agency marketing spend by 25% through the use of AI. This means reducing creative and content costs, as well as translation and other production costs. While the focus was on the removal of stock imagery, this, nonetheless, is an impressive number. 

These two press releases sent shockwaves through every CEO and CFO’s office, as they promptly walked down to the CMO and asked how their brand could also reduce their marketing budgets by 25%. As the AI revolution marches on, the question for everyone is what the impact will be on businesses and brands.

At Further, we are seeing incredible opportunities that can be unlocked by AI. However, there are some fundamental steps that need to be taken first before you can unlock this kind of savings magic. 
Here are the three things businesses need to do in order to realize these types of efficiencies with AI:

1) Education and Empowerment

Before you let the teams go crazy with AI, there needs to be some education, best practices, and governance set up.

According to Forbes, 52% of employees are already using these models, but don’t want to tell their bosses. This creates a problem for both employees and the company when it comes to these core questions:

  • How should teams be using client and company data with these models?
  • How should they treat copyright with content outputs from the models?
  • What are the best practices to share and build use cases to improve work?
  • And, for some, how do I even log in and learn to use these models?

These questions need to be answered before something goes wrong.

This doesn’t mean that the corporate overlords come in and create red tape for the sake of it, but it does mean that there are clear guidelines so that employees and brands can feel like they are ready to start to build their AI future.

2) Data Readiness

Guide to evaluating your organization for AI readiness

Your internal data structure needs to be in good shape to fully take advantage of AI. This follows the classic principle of garbage in, garbage out.

An MIT survey found that 30% of the respondents ranked the IT attributes at their companies as conducive to a rapid adoption of generative AI. Too many businesses have data scattered all over the place, and even then, that data isn’t in a great, structured shape to take action on.

Date readiness, therefore, is an important step to ensure prompts can be used to deliver high-quality and trusted outputs from the models.

3) Prototyping

Too often, we see businesses not thinking through various scenarios that might prevent something from being a successful use case for AI.

A thoughtful approach is creating a 2x2 matrix that looks at value and effort to scope out the most impactful opportunities. Once you have identified your starting point, creating some prototypes can help ensure that the desired outputs can be achieved. Don’t get so far down the path that your sunk costs don’t let you see the forest from the trees.

In summary, AI offers incredible opportunities. We are excited to see Klarna leading these efforts. However, there are clear steps that businesses need to take to fully capitalize on the promises of AI.

Learn more about Further’s Data, Cloud, and AI solutions. Contact us to take your company further.

Jason Tabeling
Head of Solutions

Read More Insights From Our Team

View All

Take your company further. Unlock the power of data-driven decisions.

Go Further Today