Having successful startups is great for an ecosystem as it leads to more startups being founded when employees (and sometimes founders) leave the original startup. Spotify and Klarna are two of the biggest ‘founder factories’ in Europe. The full report on Dealroom.co.
Matthew Ball: Big Tech’s Biggest Bets (Or What It Takes to Build a Billion User Platform). Very long and good read on how Apple, Meta, Amazon, Alphabet and others are trying or have built very large platforms. Virtual/mixed reality, gaming, data centers/cloud, iOS/Android and other examples.
Interview with Hampus Jakobsson (Pale Blue Dot)
MCJ Collective interviews Hampus Jakobsson of Pale Blue Dot, a Malmö-based venture capital fund investing in climate tech companies. The interview covers starting Pale Blue Dot and many technical aspects of venture investing raising capital for Pale Blue Dot from limited partners, what level of ownership to target in a startup, reserves for follow-on investment, how to make decisions, a bunch of things around investing in climate technology startups and much more.
Interesting both for investors who want to get insight into a peer and for founders who want to better understand how venture capitalists think.
Yesterday’s winners were not tomorrow’s winners
An interesting example from Morningstar that investing in the top 10 US companies in 1986 didn’t beat the stock market. This is another example of the historical difficulty of the top 10 companies by valuation to remain in the top 10 over 10 or 20 years.
Given the popularity and strength of the big technology companies today, it is difficult see what would happen to make them less valuable. (Not least as a capex heavy cycle in cloud have helped the largest companies, and that could happen again with AI). But historically there has always been change at the top.
Klarna on path to profitability, 1.3 billion SEK to go
Klarna reported its financial result for Q1’23. Revenue reached 4.9 billion SEK (+13 % year-over-year) and the net result was -1.3 billion SEK (-51 %).
With an overall credit losse ratio at 0.37 %, it seems like the “New geographies will behave like Mature geographies” is playing out and the overall story towards profitability is holding up, even if seems like the company will be profitable on a monthly basis earliest in the fall/early winter and not “by summer” (which I read as no later than August) that has been communicated earlier.
To become profitable by summer it seems like revenue growth needs to accelerate or other costs (mainly personell cost) will have to come down. But with an improving business and a better financing market, a few months here or there likely doesn’t matter to Klarna.
Benchmark data: SaaS startups should be in top 10 % to raise multiple rounds of venture capital
ChartMogul has prepared their 2nd SaaS Benchmark Report and it is a good, detailed read based on data from 2,100+ companies. There’s a preview on Twitter by Nick Franklin, and the full report requires registration to be downloaded.
One takeaway is that for a SaaS startup to successfully raise multiple rounds of venture capital it should have best-in-class in ARR growth rate and best-in-class retention.
Even if a startup is top quartile, it will have to think hard about if other forms of financing (bootstrapping, growth equity, private equity, becoming profitable or debt funding) are better or more likely options than venture capital, and adapt the plan accordingly.
Software spend momentum lost due to cost savings
The more cost conscious environment for software can be seen in Snowflake missing expectations. The miss mainly as companies are taking actions to lower costs. Jamin Ball writes about optimizations that lower cost for customers and revenue for Snowflake: “1) Hyperscalers releasing hardware improvements (ie Gravitron). 2) Snowflake releases optimizations to their own product. 3) Customer optimizations (like data retention). First 2 represent a ~5% / year headwind they bake in every year”
Video games is one of the strong technology sectors in the Nordics (King.com, Supercell, Paradox Interactive, DICE, Embracer and many more companies).
It’s noteworthy that Embracer, the largest publicly-traded publisher and developer, missed future revenue expectations and lost 44 % of its market value and is now valued at about 28.5 billion SEK. The reason was that a strategic partnership didn’t happen which meant Embracer missed expectations.
Embracer has had a great, acquisition-driven growth run over the last couple of years. But games development is a volatile and somewhat unpredictable business as you cannot know if the games that you are building will be fun and as a result sell well. This have been an increasing challenge for Embracer, as I understand it, as they have gone upwards from lower cost games and remakes to AA and even AAA titles with larger development budgets.
I haven’t deep-dived into Embracer’s problems, but hope they are mainly about missing expectations and that problems with revenue growth will be short-lived and turned around quickly. Not least due to the fact that a strong public Swedish games publisher is a good thing for the Nordic ecosystem.
The practical inclusion of generative AI into Adobe’s Photoshop (even if still in beta) just feels so obvious, effortless and beautiful. And Adobe being able to say that the use of the training data is not violating copyright I believe is an example of something that will be a generally important principle for AI driven services (more at The Verge).
AI model ‘speaks’ 1000+ languages based on limited training data
MIT Technology Review: Meta’s new AI models can recognize and produce speech for more than 1,000 languages. The generally more interesting point, to me, was that they were able to train the model on “two new data sets: one that contains audio recordings of the New Testament Bible and its corresponding text taken from the internet in 1,107 languages, and another containing unlabeled New Testament audio recordings in 3,809 languages”.
Open source models that can perform well using limited amounts of (internal) data seems to be a good fit for enterprise use.