Solving the Innovator's Dilemma with Gnosis (Part II)
Rihanna. One Hit Wonders. Consulting Interview Books. Innovation Models and the remarkable success of the Gnosis Hit Machine.
Thanks for reading Charterless! Subscribe for free to receive new posts and support my work.
When Rihanna auditioned for Jay-Z’s record label she performed Pon de Replay. Jay was impressed with the song, but unsure about the artist. Did Rihanna have more tracks in her or would she be a one-hit-wonder?
“I don’t sign songs. I sign artists,” Jay said.
Rihanna proved that she was, in fact, a generational talent. Jay signed her and the rest was history.
In tech, we fear one-hit-wonders almost as much as the music industry. But they happen all the time. Most companies are lucky to find a single hit. If they do stumble onto a hit, companies pivot their entire organization to support it.
This is how a gaming company became Slack.
It's how a carpooling app became Lyft.
It's how an online snowboarding store became Shopify.
Why is it so hard for a company with a hit (and the resources that provides) to strike gold a second time?
As Clayton Christensen explained in The Innovator's Dilemma: a firm with a successful product is incentivized to optimize their existing product. This requires deeply engaging with a single set of customers. It usually entails ignoring markets that you're not serving.
I was a PM in two different “innovation” product groups at Facebook. The groups were careful to follow Christensen’s advice in focusing on under-served audiences. They were also staffed with incredibly bright, creative people.
But they ran into other challenges. Employees were complacent. It was common to hear about someone withholding their best ideas for “after they left Facebook.” Bureaucratic bloat slowed down time to market. And the few truly visionary ideas that I did see were often blocked by one of the fifteen middle manager gatekeepers who “didn’t get it.”
There’s a reason that even the greats (Google, Facebook, Apple, Amazon, Microsoft) usually acquire their second hit product.
Which makes the case of Gnosis so puzzling.
As I covered in last week's piece, Gnosis began life as a prediction market. But like Slack, Lyft and Shopify, it found product market fit for an adjacent utility that it had built. The Gnosis Safe -- a wallet that requires multiple users to approve transactions — is now an industry standard. Today it secures over $107B in Ethereum assets. I can guarantee that every organization I have written about in Charterless uses it.
But Gnosis did not become Slack. They did not pivot their company around the Safe. Instead they put their head down and continued launching hit after hit.
Today, Gnosis Safe secures $107B in Ethereum assets. Its CowSwap protocol has handled 8.7B in exchanges in under a year. Its Zodiac product is becoming a standard for DAO governance. Its Gnosis Chain (formally launched in November 2021) has over 200M in locked value.
How can one five year old company achieve so much? What can we learn about innovation from how the Gnosis team approaches it?
As we’ll see the playbook likes this:
1. Craft grand vision to rally a community.
2. Break that vision into constituent, atomic products.
3. Incubate or fund fully autonomous teams to build those atomic products.
4. Spin them out as independent organizations.
Let’s see it in action.
The Grand Vision
In 2015, Martin Köppelmann and Stefan George started Gnosis as a project of Consensys. The pair had collaborated on a Bitcoin sports betting market in 2013. But with Gnosis, they set their sights on a bigger prize: the world’s first decentralized prediction market.
There was just one big problem.
None of the technology that they would need to build Gnosis existed yet.
But that challenge presented its own opportunities.
The path to building a great product is often to solve your own problem. So that’s exactly what Gnosis did. Repeatedly.
Gnosis Safe - An Atomic Product
Multisignature Wallets (multisigs) require multiple users to consent before completing any transaction. They have been around since the early days of Ethereum. But the original multisigs only supported transactions in Ether.
The Gnosis team was developing two new tokens of their own, GNO and OWL. The Gnosis team knew that they would need a new multisig to store their tokens. So they built one. Then they released it to the broader community.
Their timing could not have been better. In 2017, team after team launched their new tokens on Ethereum during the ICO boom. By August 2017, their Multisig was already securing $2B in tokens.
Gnosis had found something any startup founder would kill for: clear product market fit in a fast-growing market.
Then crypto winter hit and the whole thing almost came crashing down.
This turned out to be a fortuitous break for Gnosis. Rather than pivoting their team, they had time to recommit to their original vision. Their vision was bigger than a single product. It would require several more big product wins to realize.
In a 2018 letter, their CEO Martin Köppelmann outlined their strategy and the products they were developing to achieve it.
1. Allow anyone to create tokenized prediction markets. This project would become their first decentralized prediction market: Omen.
2. Allow anyone to trade these tokens easily. To achieve this they built both DutchX and Gnosis Protocol, now known as CowSwap.
3. Continue to invest in Safe to hold tokens for prediction market users.
Gnosis could have chosen to build these features as an all-in-one suite. They were, after all, far ahead of their competitors. The dominant strategy might have been to consolidate their gains.
But instead, the Gnosis team took a page from the Amazon Web Services playbook. Instead of integrating, Gnosis built each tool as a standalone product. They could be used by any team for any purpose.
Köppelmann saw that these tools could form the backbone of Ethereum's ecosystem. When crypto winter turned to spring, that’s exactly what happened.
Beyond Build, Buy or Partner: Incubate and Exit
Everything I know about business I learned from the consulting interview prep book, Case In Point. It teaches that a company entering a new market has three options: build a new entry, buy an existing player or partner with an existing player.
Surveying the work required to realize their vision, the Gnosis team faced a similar question. Should they build every possible tool in-house? Should they buy small startups building useful tools? Or, should they partner with startups to launch co-brand products?
Instead, Gnosis threw out the playbook. They decided on two new strategies: fund and incubate/exit.
The “fund” approach is common in web3. Many projects create ecosystem "grant funds" subsidize useful projects. But the incubate and exit approach is unique to Gnosis.
Gnosis’s prediction market utilizes a unique token for each bet. Traders can buy or sell these tokens as they please. But a token without any potential buyers is useless. So Gnosis needed a way to ensure that all of these tokens could be easily traded.
They built CowSwap to solve this problem. CowSwap uses an approach called “ring trades.”
Imagine that I have Token A and want to buy Token B. John has Token B, but wants to buy Token C. Sarah has Token C, but wants to buy Token B. In a traditional swap, it would be impossible for us to trade. Ring trades would match our "coincidence of wants" so that we could profitably trade with each other. This increases the tradeability of rare tokens while keeping transaction fees low.
The Gnosis team believed that CowSwap could be used to optimize every major DEX in the Gnosis ecosystem. But how could they achieve that kind of growth for their new venture? Köppelmann proposed that Gnosis spin out the CowSwap project as its own DAO.
This would be like Google spinning out Gmail in 2005. To put it mildly, it’s not how things are usually done. There are traditionally advantages to keeping separate product efforts together. The company can benefit from pooling talent, shared infrastructure and shared distribution.
But in a decentralized ecosystem, these advantages breakdown.
Talent flows flexibly across projects so Gnosis and CowSwap do not need to worry about aggregating employees. The infrastructure built by Gnosis is open source so there is no advantage to building “inside” the organization. And Gnosis can continue to distribute CowSwap.
In contrast, the advantages of decoupling in web3 are much clearer.
This move enables CowSwap to issue its own tokens in line with its own goals. CowSwap would not need to worry about second-order effects on Gnosis. Issuing tokens to encourage users and investors is essential for Web3 projects. From the outside, we all see the speculative frenzy it creates. From the inside, it bootstraps new ecosystems at lightning speeds.
At the time of writing, CowSwap represented a relatively small share of Gnosis’s value. As a more mature project, Gnosis’s upside is also relatively more constrained than a new, riskier, but potentially higher-growth bet. To encourage investment from new participants, CowSwap needs the flexibility to offer tokens with a much more aggressive upside.
In exchange for incubating CowSwap, Gnosis received 10% in the new CowSwap project with the development team receiving 15%. That represents considerable dilution. So the risk Gnosis is taking is clear. The DAO is betting that the gains from independent decoupled execution and tokens will compensate for dilution.
This may be a risky bet. Spinoff stocks have historically outperformed the S&P, but this trend has reversed in recent years. Gnosis, however, isn’t waiting to see if it pays off before doubling down. It has since announced plans to spin out Gnosis Safe. We can expect that Chain and Zodiac will follow suit, too. If Gnosis succeeds they will flip the conventional wisdom about The Innovator's Dilemma. Gnosis may not only build one generational company. They may just build five.
Or check out some open roles with the team:
Business and Ecosystem Development (m/w/d)
Remote, Berlin, Gibraltar