social networks in Europe
#144
Howdy folks, welcome to Sunday CET. This week:
social networks in Europe
a new bubble is brewing
will (Open) Ai kill Google
Enjoy,
Dragos
Observations
European social networks
Europe was never a proper environment for building huge social networks. It is a super fragmented market, w/ 40+ countries and 24 official languages making it difficult to grow a tech product fitting an universal user behavior pattern (read time consuming and expensive). That is a reason for which companies such as Google, Facebook, Linkedin or Tiktok had it easier to enter and get significant market share in Europe compared to the local incumbents - they already had mature products, with proven product market fit and established GTM blueprints, and adapting to European markets was way faster and cheaper. And so, with each tech cycle, from 2000s onwards, American companies building solid network-based communities took the lion share and startups such as Dailymotion or Jaiku couldn’t compete vs Youtube or Twitter respectively, for example.
Back to present day, the more remarkable European social network story is Bereal, which in 2022 alone grew 100X, from 1 million to 100 million downloads and 20M MAUs, mostly American. It is the straight-to-US GTM playbook that led to this traction in a mere 2 years of existence - compared to another French social network, Zenly, which got to 40 million MAUs in 12 years, 5 of which under Snapchat’s management. And, as a side note, Zenly was sold because the rate of user growth was slower than than the spending rate for getting that growth, and so the VCs where at an inflection point where it was more profitable to extract value from what they thought to be at peak, rather than throwing more money in order to wait for a future better liquidity event. Keep in mind that the VCs judge their assets in 5-8 years slots, as they have to return money to their LPs - good business requires understanding and gaming the odds of cashing out vs waiting for a multibillion valuation in that timeframe. That is also why they push startups for fast growth - their limited time window becomes the startup’s problem to grow as fast as possible, that’s how the VC business works.
All this context as this past weeks, we’ve had a few announcements of new social networks in the making in Europe - Amo, created by Zenly’s founders and trying to prove Evan Spiegel wrong with closing it down and Slay, a Gas clone made by three German kids, already pre-seeded by VCs due to some decent local traction (Swedes got a local clone too, btw, but weak traction). I crunched out the Nordic 9 data and found quite a few other VC-backed startups building social networks - 82 investments in 2022 and and 94 in 2021 - early stage only (<$5M), that is. Notable examples include:
business: Melon, We Don’t Have Time, Equel Social, Zeed + Polywork which raised $16.5 million already.
music: Logcast, Keakr, Colkie, KaraokeOne, Frow
metaverse: Cardalonia, Challau
A new bubble is brewing
Every VC and their mother is excited about OpenAI/ChatGPT in particular, and startups using this piece of tech for building stuff out in general. It’s a legit good thing, at least this narrative feels different and more interesting than the endless talk about web3 saving the world, crypto scams and the VC bubble deflating, which kept the 2022 agenda. And that’s great - there is a whole class of venture capitalists hopefully also getting deeper into the difference between AI and ML and what it takes to actually collect proper data and train models in order to produce a commercial product.
I can’t help but notice though that all this sudden VC emotion about AI-based projects is actually about building stuff for a single platform operated by a VC-backed startup (OpenAI), with a generic product appealing to consumers and businesses altogether. Excitement aside, if you build your stack and a business on top of it, you’re locked in at a competitive dis-advantage - and onboarding in as many startups using their code is exactly OpenAI’s strategy here to become a significant operation.
Investors backing platform-dependent ecosystems is not exactly new in the contemporary tech startups history - I am old enough to remember founders gaming pitches by inserting “the Facebook/Twitter/4Square play strategy” in their decks, because at that time those were the new shiny platforms everybody was excited about. Right now, that ploy translates into adding an “AI-based” at the front of a startup’s name, justified by paying 20 bucks a month’s ChatGPT subscription - “AI-based trading co”, “AI-based research co”, “AI-based FBA aggregators”, “AI-based 15 minute food delivery” etc. And my favourite: “AI-based VC”. This fad shall pass as well, it’s like calling a startup a mobile company just because it makes money from a mobile app.
OpenAI is just history repeating, and basic economic theory tells us that when you have one sole tech stack provider and a lot of demand to build on top of that from the investor side, the prices will go up quickly and we will get to a VC bubble again. Oh wait, prices for anything “AI-based” are already up, aren’t they?
Not least important, if you build on Open AI’s rails the so-called AI is not your IP, but rather what you build on top of it, hopefully with a smart biz model - yet that is a very risky business move, because the AI supplier can change rules and costs anytime they wish, just like Facebook has been doing with their platform, for example. With Google and Meta working on similar AI-as-a-service platforms, we may get to a category that should become a somewhat reliable and predictable business environment.
Good investors of course will look for differentiated startups building their own data sets and train their own models for specific business cases. But that can’t be done over night, it’s costly and actually requires heavy work of data science and engineering, unlike deploying on the top of an off-the-shelf product that does generic stuff. The good news is that, unlike past years, now it is easier to sell AI-based assets to investors, and that’s a gain for engineers solving hard problems - the funding pool is larger and the familiarity is up a notch on the buy side.
Will (Open) AI kill Google?
That’s just a silly narrative I keep seeing around - people are quick to dismiss Google just because it screwed a rushed demo this week and all this led to a short-sighted market sentiment. However, that sentiment rather reflects the feelings for Google in general (Google is not exactly a love brand) combined with rooting for a lousy incumbent.
Rationally speaking Google’s business and products are still super solid. Bing is an inferior product with an inferior business model, and plastering ChatGPT on top of it won’t make it better over night, let’s just get real here. Secondly, don’t underestimate Google, in spite being an evil monopolistic company, it employs some smart people who do hard engineering work - that’s not advertised because their business model is selling ads, not tech, and not all analysts understand its modus operandi. Yes, the management may be bad, but Google’s business model is very hard to break as it incorporates many competitive advantages built in time and reinforcing each other, and implicitly it will take a long time until a competitor will steal their thunder. Taking the over is still on the government breaking it rather than Bing kicking its ass. Or ChatGPT for that matter.
The January numbers
Here’s the January 2023 numbers for the Euro VC data we have tracked over at N9 so far*:
Startup investment deals:
372 deals | $3.5bn value | 43 with undisclosed data
vs.
612 deals | $12.3bn value | 82 with undisclosed data [January 2022]Active investors in January 2023 - 961:
VCs: 🇩🇪 HTGF (6 deals), 🇬🇧 Elbow Beach (5 deals), 🇧🇪 Finnidus (5 deals) - see all
angels: 🇳🇴 TRK (2 deals), 🇮🇹 Davide Dattoli (2), 🇮🇪 Brian O'Hagan (2) - see all
Investment distribution by region:
UK - 87 deals
Nordics+Baltics** - 70 deals
France - 66 deals
DACH - 66 deals
Southern Europe***: 40 deals
Benelux: 26 deals
CEE****: 18 deals
🇬🇧 London 58 deals
🇫🇷 Paris 37 deals
🇸🇪 Stockholm 20 deals
🇳🇱 Amsterdam 15 deals
🇩🇪 Berlin 15 dealsInvestment distribution by size/value:
< $0.5M - 26 transactions (7%) / $9 million (0.3%)
$0.5M - 1M - 33 transactions (9%) / $23 million (0.6%)
$1M - 5M - 151 transactions (41%) / $378 million (10.7%)
$5M - 10M - 38 transactions (10%) / $257 million (7.2%)
$10M - 20M - 45 transaction (12%) / $585 million (16.5%)
$20M - 50M - 21 transactions (6%) / $685 million (19.3%)
$50M - 100M - 8 transactions (2%) / $508 million (14.3%)
$100M+ - 7 transactions (2%) / $1.1 billion (31%)
n/a - 43 transactions (12%)
TOTAL - 372 transactions / $3.5 billion
Verticals (by number of deals)
Software 215 deals - $1.5 billion value
Manufacturing 113 deals - 1.6 billion value
Marketplace 75 deals - 628 million value
Health services 69 deals - 519 million value
Financial Services 64 deals - 646 million value
* The usual reporting lag is at 30%, equity deals only etc.
** Sweden, Norway, Finland, Denmark, Iceland, Latvia, Estonia, Lithuania
*** Italy, Portugal, Spain.
**** Bulgaria, Croatia, Czech Republic, Slovenia, Greece, Hungary, Poland, Romania, Serbia, Slovakia.
Cheat Sheets
Sign up and become a Nordic 9 subscriber from here - your support is greatly appreciated!
Good to know basis
👉 How much do investors want - no biggie
🤔 Is cold emailing working with investors? - of course
🇸🇪 Guide for startups in Sweden - basic stuff
🇩🇪 Guide for startups in Germany - basic stuff
🇪🇺 Cloud Challengers report - top 100 early stage, B2B SaaS/Cloud companies
✨ Nordic B2B Software Growth - benchmark report 2023
🕸️ The new gatekeepers - a take on the big picture
Have you produced useful resources for the community? They can be events, reports or tools - send me an email with submissions at drnovac@gmail.com
Other notes
🇫🇷 Old money is old Patriarch Bernard Arnault is ensuring the Louis Vuitton empire remains a family business, installing his kids in key roles in a new holding company, control of which is split equally among his five children - Delphine, 47, and Antoine, 45, from a first marriage, and Alexandre, 30, Frederic, 28, and Jean, 24, a trio he had with his wife, the Canadian-born pianist Helene Mercier. Dude is 73 and says he has no intention of stepping down anytime soon. European HNWI ftw.
🇪🇺 Europe is just slow and hyper-regulated, kind of like a museum or a nice place to retire. Of course - but problems can be turned into opportunities, n'est ce pas?
🇬🇧 Meanwhile in London luxury Mayfair homes are selling at the fastest rate since 2020.
Related: is $400k/yr in New York rich?
👋 Founder's choice VC leaderboard. Hoxton (#8) is the only European in top 10.
🤑 True story The venture capital industry, which desperately needs to find a new The Current Thing to get excited about, switched from the ‘web3 is the future"‘ to ‘AI is the future’ narrative. link
(Somewhat) related: Kim Kardashian paid $1 million to speak at Miami hedge fund event.
Yeah, she’s hot, but I came in from London to do business. What am I going to learn from her?
I know I am not Kim, but if you want to take advertising over at Sunday CET, my rates are lower than hers. 😛 Ping me to find out.
🦇 Americans are being ridiculous The government is holding up a proposed sale of Forbes magazine to an Indian billionaire who amassed much of his fortune in dealings with Russian state-owned enterprises. Forbes is considered a strategic asset - it used to be a good brand in the 90s, nowadays it’s just the front of a low-quality farm for sponsored-content.
📺 What will TV look like in three years? These industry insiders share their predictions
🗣️ Hey Arnold. Good interview with Arnold Schwarzenegger.
🇳🇱 Tourists. Amsterdam plans to ban smoking marijuana on the streets of its red-light district.
🎨 Los Angeles is starting to paint many streets white. The higher reflectivity would reduce city summer temperatures by up to 10 degrees Fahrenheit.
⚡ Tech disrupters, a handy guide.
❤️ Make love not war. Swedish city launches love week to battle plummeting birth rate.
🤔 Do you know how to behave? How to text, tip, ghost, host, and generally exist in polite society today.
🦍 You can run but you can’t hide Italian mafia fugitive Edgardo Greco was arrested after 16 years on the run, working at pizza restaurant. He was busted because.. wait for it - he was featured in a local newspaper in 2021, advertising the restaurant's authentic Italian cuisine.
Did you find this email useful?
Great, will forward • Good • Meh
Thanks for reading! Please send me your thoughts or comments by hitting reply.
If you would like to sponsor this newsletter, please get in touch. I am bit biased, but Sunday CET is a great way to reach investors and founders from Europe.
If this email was forwarded to you, please subscribe, it’s free!
Created every Sunday by @drnovac of Nordic 9 with weekly notes and observations from the European startup ecosystem.
You have received this email as you signed up at Sunday CET or are a Nordic 9 registered user.
Dragos, great stuff again!
I especially liked the piece on Web3 and Ai - I remember when nobody was interested in GenerativeAI and we invested in it and everybody was like „uh?“ and instead did web3, defi/crypto stuff...