Klarna is a Swedish verb that translates to ‘clearing up’ or ‘becoming clear/clearer’

Klarna filed for an IPO on March 14th, , marking the culmination of a volatile journey for the BNPL firm. The company reached a peak valuation of $45.6bn in 2021, then declined dramatically to $6.7bn in 2022. For its upcoming IPO, Klarna is targeting a valuation between $15-20bn. This offering will serve as a bellwether for market sentiment towards fintech companies, potentially opening the doors for Revolut, Chime and Stripe. A couple of notes:

AOV

aov

Klarna and Affirm show divergent AOV trends from 2022 to 2024, with Klarna increasing from $95 to $100 (2.5% CAGR) while Affirm declined from $374 to $292 (-11.6% CAGR).

Klarna’s modest AOV growth stems primarily from expansion in their U.S. business, where AOV is generally higher than in their other markets. With 69% of their U.S. purchases concentrated in apparel and accessories, Klarna has built a model centered on frequent, lower-value transactions. This approach is evident in their user patterns, with card users making approximately 126 annual purchases per active consumer, though this remains below the 251 annual purchases typical for U.S. credit cards.

Affirm’s declining AOV reflects a deliberate strategic shift rather than weakening performance. Despite initially focusing on retail e-commerce and travel sectors with naturally higher transaction values, Affirm has actively diversified their merchant base while promoting repeat platform usage beyond one-time high-value purchases. Their expansion plans targeting in-store purchases, leisure, and utilities further indicate this pivot toward frequency over individual transaction size. To their credit, Affirm’s $292 AOV remains impressive considering 80% of U.S. e-commerce transactions fall under $150.

Transaction patterns between the companies reveal deeper differences in their business models. Klarna customers maintain an average outstanding balance of just $87, compared to $6,730 for traditional credit cards, indicating fundamentally different usage patterns. Lower AOV often correlates with increased purchase frequency, particularly evident in certain verticals like Events and Services, including Transportation.

Continued improvements in credit underwriting capabilities and expanding merchant networks, both Klarna and Affirm could increase purchase frequency and expand AOV. Affirm’s strategy is one path Klarna could emulate going forward.

AI

Klarna directly attributes ~$44mn (~2% of opex) in cost savings in 2024 due to AI interventions (customer support + internal tooling). They also suggest that a good chunk of marketing cost reductions are driven by AI:

We have combined AI adoption with a rigorous approach to managing operating costs to drive significant efficiency savings. For example, our sales and marketing costs declined to $328 million in 2024 from $531 in 2022, driven by AI implementation, cost efficiencies and process centralization, while we maintained a strong GMV and revenue growth. In particular, we reduced our spending on marketing agencies by 75%, from $33 million to $8 million.

Optimistically, if we add another $44mn costs saved here, total yearly savings comes to about ~88mn, or 4% of opex, which is marginal at best. And while this is not a 1-to-1 comparison, their R&D spend increased 14% during this period, whereas customer service and operations costs decreased by 15%, nearly canceling each other out.

So it’s interesting that in a moment of narrative flourish, Klarna attributes annual revenue/employee increasing by ~140% in 2 years to AI:

AI transforms our productivity and drives increasing efficiency. AI adoption—including the related reduction in the use of third-party suppliers and vendors and the adoption of the AI copilot to create and review code—has led to internal efficiencies. Our average annual revenue per employee at period end has increased from approximately $344,000 in 2022 to approximately $821,000 in 2024.

What’s conveniently downplayed is how the macroeconomic environment massively benefited Klarna’s core business model. Interest rates shot up dramatically during this period, which means as a bank, the business of lending money became substantially more profitable almost overnight. When evaluating the true impact of AI investments at Klarna, we should be asking what portion of that 140% revenue/employee increase would have happened anyway due to the interest rate environment that has benefited all lending businesses during this period. The underlying subtext is “we want to trade at AI co. multiples and not bank-with-$9.5bn-in-deposits multiples”.

Funding Sources

Neither Klarna nor Affirm has pursued a banking license in the US market so far. Depending on the rate environment, analysts believes consumer deposits could cut Affirm’s on-balance sheet funding costs in half. Klarna has deliberately chosen to avoid deposit-taking operations in the American market, as stated in their regulatory filings:

We do not take deposits in the United States, including interest-bearing deposits, as we do not maintain the necessary banking licenses to take U.S. deposits. Given our access to a variety of funding sources, as described below, and our decreasing net losses in recent periods, expanding our deposit-taking operations into the United States is not currently a part of our funding strategy.

Klarna already has a banking license in the EU so they have less of a reason to pursue it, compared to Affirm. But consumer deposits will be cheaper still to finance loans? Maybe regulatory oversight isn’t worth the bps.