Father, Husband, Network Effects Nerd & Partner @b2venture | Alumni: @Speedinvest @Uniplaces @PointNineCap @Naspers @ricardo_ch
Berlin
Joined November 2014
- Tweets3.1K
- Following2.3K
- Followers4.5K
- Likes6.8K
Pinned Tweet
1/2 I first backed @BergstromTommi and Elmo in Ambio's pre-seed round. Today, @b2venture is leading their €5.1m seed round alongside Montis and @LifelineVC.
Very happy to back them again as they make recycled plastics easier to buy!
1/4 Some incumbents can afford to let AI agents own the customer interface more easily than others.
That may create room for new headless marketplaces: businesses that aggregate supply and enable transactions behind agents' interfaces.
3/4 @danhockenmaier argues that @amazon has little reason to open up before agents bring meaningful demand.
That delay could create a window of opportunity for new headless marketplaces to build supply and serve demand before incumbents strike deals.
4/4 The window is only useful if entrants build something durable. Trust and fulfilment still have to be delivered.
New entrants need to make that work; however, without an existing profit pool to protect.
#headless #headlessmarketplaces #marketplaces
danhock.co/p/how-scared-shou…
1/3 A very late Sunday thought: I think @usv 's and @benchmark's larger early-stage funds are partly playing catch-up with valuation inflation.
2016-2026:
US Seed median pre-money: +290% (H1)
USV Core: +186%
Benchmark flagship: +76%
@PitchBook /@nvca: nvca.org/wp-content/uploads/…
2/3 When buying a comparable stake costs more, a larger fund can help preserve a disciplined investment strategy.
@usv describes roughly the same investment count, but bigger rounds, higher prices and longer support.
USV's explanation: usv.com/writing/the-next-cha…
3/3 Exit values matter too. @SpaceX's ~$1.8tn IPO valuation shows how large potential outcomes can now be.
The fund-size debate should account for both higher entry prices and the scale of possible exits.
A larger fund can be a way to preserve a disciplined investment strategy.
Same pattern in 🇩🇪: €4.5bn invested in Q3, €3.5bn+ in just six €100m+ rounds, while €1m+ rounds fell to 89.
My concern: less capital at discovery stage may mean fewer companies get the chance to become future large rounds.
We are in the age of consensus capital [2026 part II]:
1– Only 10% of LP funding to emerging managers
2– Only 10% of VC rounds smaller than $50M
@adjacent is for the unproven
I went a bit deeper on the German data and why I think this matters for the future growth pipeline here:
lnkd.in/p/dAypUVkq
2/2 Ambio fits my HAHO thesis: AI can scale complex trade by automating coordination. Its merchant-of-record model adds accountability for quality and delivery.
That's why I'm bullish on AI-native B2B marketplaces.
b2venture.vc/stories/why-we-…
Mathias Ockenfels retweeted
Technical moats are mostly over at this point. If you’re building something that is meant to be durable, the only moats now are:
• Proprietary data
• Network effects
• A physical component
1/3 #Instinct will die Angebotsseite zur Kasse bitten: eine Take Rate für vermittelte Transaktionen. Als Marktplatz-Investor finde ich das spannender als die nächste Bewertungsrunde. Wie viel Provision kann ein AI-Agent tatsächlich durchsetzen?
2/3 Für mich hängt das an der Verhandlungsmacht: Wie wichtig ist der Anbieter für den Agenten, wie abhängig ist der Anbieter von dessen Nachfrage? Eine angekündigte Take Rate sagt noch wenig darüber aus, was Anbieter am Ende zu zahlen bereit sind.
3/3 In der neuen Folge von @StartupInsid_er Investments & Exits spreche ich mit Jan Thomas über Instinct, die Take Rate und darüber, welche meiner Einschätzungen aus der letzten Folge ich inzwischen korrigieren würde.
open.spotify.com/episode/5b2…
I am investing post AI.
When everything is AI, nothing is AI.
A model lab, an accounting product and a marketplace can all carry the label. Yet their economics and competitive dynamics are worlds apart.
As AI becomes the default, the investment questions get more specific:
“Why the Germans Do It Better – Notes from a Grown-Up Country.”
Spotted in Berlin-Mitte today. Left on the sidewalk, free to take, between old shoes and discarded household stuff.
Disclaimer: haven’t read the book.
I’ll resist making the obvious joke.
1/3 Weekend reflection: AI may commoditise products. But network effects may be one of the few moats it reinforces.
@nvidia x @huggingface and @Stripe x @OpenRouter were acquired at ca. 80-85x and 50-60x rev. That is a strategic premium, not normal vertical AI/software pricing.
Sources:
Hugging Face × NVIDIA
reuters.com/business/nvidia-…
OpenRouter × Stripe
reuters.com/technology/payme…
