Thread: “Fake it til you make it” is a key strategy for overcoming the chicken-and-egg problem of starting a new marketplace, popularized by companies like Uber and Yelp. The idea is to bootstrap one side of the marketplace inorganically in order to attract the other side.
There are a few tactics that fall under this umbrella of faking it in a marketplace--including paying guarantees or subsidizing transactions, managing supply, or producing the supply/demand yourself. Then, at a certain point, network effects kick in and organic growth takes over.
A few examples: 1) Uber launched by going to black car companies and paying drivers to be available on Uber during certain hours, ensuring that riders would be able to find a ride.
2) Relationship Hero, a relationship coaching marketplace, scaled to dozens of customers with just one coach--its cofounder! But the website listed 10 fake coaches, to give users the sense that it was a more active platform with diverse coaches who fit their particular situation.
(Today, all the coaches listed on Relationship Hero are real)
3) Beepi, which was a used car marketplace, had a massive chicken and egg problem in attracting sellers and buyers initially. To solve this, the founders went out and purchased used cars to seed the supply side. After a few months, they moved to the marketplace model.
4) Managed marketplaces are also a form of "fake it til you make it," where the supply side is employed or otherwise managed by the company. This model is frequently applied to complex services, to create a radical improvement in the user experience.
Sometimes these managed marketplaces preserve the managed aspect while scaling, while others open up more broadly (e.g. the Bird platform).
"Fake it til you make it" can be applied not just to building marketplaces, but all sorts of new networks, like dating, social, etc. Creating new marketplaces from scratch and overcoming the cold start is challenging. We love seeing creative strategies here.
My partner @cdixon has also written about the “come for the tools, stay for the network” approach. Let us know if any others come to mind! cdixon.org/2015/01/31/com…
• • •
Missing some Tweet in this thread? You can try to
force a refresh
In my last tweetstorm on crypto and AI’s intersection, I described projects that are applying crypto to solve AI problems.
Today, I’ll outline the second category: companies that are applying AI to crypto.
AI's application to crypto parallels the broader impact we’re already seeing AI and LLMs have on the non-crypto world:
Creating a new user interface for computing; helping content creators in their creative processes; helping enterprises enhance their operations
The category of projects applying AI to crypto can be segmented by GTM: consumer, creator, or developer-facing.
On the consumer side, projects are building:
- AI interfaces for users (e.g. a natural language interface for wallets to do things onchain)
- AI trading agents and strategies
- Valuation models (e.g. Upshot with NFTs)
- User-owned AI DAOs (e.g. @hollyplusp_ voice model, @BottoDAO)
- Onchain AI games (e.g. @ModulusLabs’ Leela)
I've been spending a lot of time at the intersection of AI and crypto. We believe the coupling of the two is critical to making AI more equitable and user-owned.
Sharing quick thoughts about what we’re seeing:
Projects at the intersection of AI x crypto fall into two broad categories:
1) Applying crypto to solve AI problems 2) Applying AI to solve crypto problems
Because the market for AI overall is larger today than the world of crypto, most companies are focused on (1) applying crypto to AI.
That said, applying AI to crypto (2) can help expand the pie by making crypto easier to use and interact with.
A metric that I like to ground convos about crypto in is that Ethereum just crossed 245m cumulative unique addresses. Meanwhile, the internet had 248m users in December 1999.
We're at December 1999 levels in terms of onchain activity
December '99 was just months before the dot-com bubble burst, and many well-funded, hyped businesses failed and shut down
Many of those businesses actually had great ideas, but were just too early: drkoop -> WebMD; Webvan -> Instacart; -> ChewyPets.com
1999 was also years before we got the major companies of the web2 era -- FB in 2004, MySpace in 2003, YT in 2005 -- that gave users lots of reasons to spend time online
Psychological ownership is the feeling of possession or "mineness" over a product/service. It’s distinct from legal ownership: you can *feel* like an owner without actually owning something (my sports team, my social media profile), and vice versa.
Psychological ownership is important for product builders because it changes behavior. It can increase loyalty, word-of-mouth growth, and willingness to pay. In digital communities, feeling ownership leads to increases in satisfaction, self-esteem, and contribution quality.
Lots of thoughts about the TikTok hearing last week:
- it's horrifying to see the level of knowledge that lawmakers have about the internet/tech ("Does TikTok access users' home wifi networks?" was a real q that was asked)
- a ban of TikTok will leave 150m users orphaned, as well as 5m SMBs that use it to grow organically
- less competition among social media companies -> worse for creators & users. e.g. TikTok pioneered the Creator Fund and sparked a slew of competitors (Reels, Snap, Pinterest)
- users spend on average 95 mins/day on TT. that's a huge vacuum of time. lots of opportunity for new social media entrants
- TikTok's algo provided "universal basic distribution," or the chance for any video to be discovered, which was much harder on legacy social media
We’re excited to announce the inaugural Variant Founder Fellowship: a three-month cohort-based program for web3 founders at the earliest stages of their journey:
Variant is proud to support some of the top projects in the ecosystem, which are part of our Variant Network, a peer learning community comprised of the founders and leaders across our portfolio.