I love a good scoreboard, not only because I enjoy competing, but also because it reveals a great deal about the people it is scoring. Crypto has had many scoreboards over the years, some of which I've spoken about. I won't rehash all the historical details, but let's revisit one example: the L2beat TVL scoreboard. For several years, this was THE determining factor for which Layer 2 was "winning."

Why was this the scoreboard? Total Value Locked (TVL) is a somewhat nonsensical metric for DeFi; it's even less relevant to L2 solutions. However, it is measurable and relatively difficult to game. Unfortunately, it exemplified the principle that "you only improve what you measure," and its corollary, "you should only measure what you want to improve." In the last cycle, we saw multiple L2s aggressively gaming this scoreboard. Well played, I say.

However, during this cycle, a new scoreboard has arisen: Kaito. Unlike previous metrics that served as proxies for attention, Kaito directly measures attention and mindshare. Kaito didn't invent “mindshare", but it successfully quantified it. The brilliance here is that attention is exactly what every other scoreboard has been indirectly tracking.

Now that we have a quantified attention metric at both project and user levels, numerous applications are possible. Historically, TVL campaigns signalled safety; if SBF dumped $100m into an LP, it seemed safer. Alameda surely did some technical due diligence… Although that may not have been the signal we thought it was. Over time, through the Lindy effect, TVL became a more accurate proxy for safety. But interestingly, TVL also evolved into a second-order proxy for attention and mindshare. Greater TVL implied greater interest, but while directionally correct, it remained highly susceptible to gaming. Eventually, TVL declined in power, prompting pivots toward metrics like fees, revenue, and transactions per second (TPS).

I've jokingly suggested Infinex should introduce a metric called "Signups Per Second"—our peak currently sits at 33. I'm sure Moonshot has us beat, though...

Instead, we've decided to embrace this new attention-driven metric. Our campaign specifically rewards those who consistently generate mindshare. Why?

Our thesis is simple: mindshare drives signups. And high-quality mindshare drives high-quality users. Direct attribution, like referral codes, is useful, but many users deliberately or unintentionally avoid using them. Still, someone influences these users to sign up. We believe that at least 50% of signups result from general mindshare. A concept we term "indirect attribution."

Daily, hundreds of tweets mention Infinex, but only 5-10 achieve significant reach. Those tweets likely account for a meaningful percentage of new signups. Thus, we attribute subsequent signups to those who dominate mindshare about Infinex on X each day.

This means that every signup on Infinex during Yaprun is attributed to a cloud of X accounts inside the Infinex mindshare cloud.

In addition to indirect attribution, we still offer referral codes. If you convince everyone in your private Telegram group to sign up, you'll receive 50% of the benefit; the other 50% goes to the individuals who dominate that day's mindshare. While seemingly counterintuitive, signups rarely derive solely from referral codes. Typically, a user's intention to sign up comes from multiple interactions and sources.

This novel campaign isn't about driving superficial attention; it aims to generate meaningful awareness for Infinex. We call this approach "Mindshare Farming."

Kaito has recently received criticism on the timeline, criticism that's partly justified, as campaigns rewarding low-quality AI-generated content harm the ecosystem. That's why we've developed a proprietary algorithm to carefully measure content quality specifically for Infinex, overlaying this quality metric onto Kaito’s raw data. While Kaito has attracted some top Twitter accounts, many have remained cautious. We believe our campaign will clearly demonstrate that high-quality content about Infinex is rewarded, and low-quality "slop" is discarded.

Of course, botting and sybil attacks are risks for any campaign. Fortunately, Kaito’s "smart follower" metric is costly to manipulate, leading to higher account and content quality. Twitter's monetisation system already demonstrates the value of high-visibility accounts; our campaign amplifies this effect tenfold. With approximately $6 million USD distributed over four seasons, I expect this to usher in a new meta and that other projects will deploy similar campaigns.

A minor note: teams are often excluded from campaigns like this. While I disqualified myself and redistributed the rewards to everyone else on the leaderboard, I felt strongly that the rest of the team should also get to participate. Having a wide variety of perspectives across the team that are actively engaged on the timeline has been a significant boost to engagement over the last year. I encourage founders to let their teams Yap.

One final point: retroactive rewards are a powerful way to prevent gaming, so that's what we opted for in Season Zero, which concluded in stealth this week. There were numerous hints and alpha drops over the last few weeks as we finalised the campaign details, but we didn’t actually freeze the scoring algorithm until the absolute last minute. If you were paying attention, you might find yourself here.

For full details of the campaign, please refer to this post.

Yapfinex

I love a good scoreboard, not only because I enjoy competing, but also because it reveals a great deal about the people it is scoring. Crypto has had many scoreboards over the years, some of which I've spoken about. I won't rehash all the historical details, but let's revisit one example: the L2beat TVL scoreboard. For several years, this was THE determining factor for which Layer 2 was "winning."

Why was this the scoreboard? Total Value Locked (TVL) is a somewhat nonsensical metric for DeFi; it's even less relevant to L2 solutions. However, it is measurable and relatively difficult to game. Unfortunately, it exemplified the principle that "you only improve what you measure," and its corollary, "you should only measure what you want to improve." In the last cycle, we saw multiple L2s aggressively gaming this scoreboard. Well played, I say.

However, during this cycle, a new scoreboard has arisen: Kaito. Unlike previous metrics that served as proxies for attention, Kaito directly measures attention and mindshare. Kaito didn't invent “mindshare", but it successfully quantified it. The brilliance here is that attention is exactly what every other scoreboard has been indirectly tracking.

Now that we have a quantified attention metric at both project and user levels, numerous applications are possible. Historically, TVL campaigns signalled safety; if SBF dumped $100m into an LP, it seemed safer. Alameda surely did some technical due diligence… Although that may not have been the signal we thought it was. Over time, through the Lindy effect, TVL became a more accurate proxy for safety. But interestingly, TVL also evolved into a second-order proxy for attention and mindshare. Greater TVL implied greater interest, but while directionally correct, it remained highly susceptible to gaming. Eventually, TVL declined in power, prompting pivots toward metrics like fees, revenue, and transactions per second (TPS).

I've jokingly suggested Infinex should introduce a metric called "Signups Per Second"—our peak currently sits at 33. I'm sure Moonshot has us beat, though...

Instead, we've decided to embrace this new attention-driven metric. Our campaign specifically rewards those who consistently generate mindshare. Why?

Our thesis is simple: mindshare drives signups. And high-quality mindshare drives high-quality users. Direct attribution, like referral codes, is useful, but many users deliberately or unintentionally avoid using them. Still, someone influences these users to sign up. We believe that at least 50% of signups result from general mindshare. A concept we term "indirect attribution."

Daily, hundreds of tweets mention Infinex, but only 5-10 achieve significant reach. Those tweets likely account for a meaningful percentage of new signups. Thus, we attribute subsequent signups to those who dominate mindshare about Infinex on X each day.

This means that every signup on Infinex during Yaprun is attributed to a cloud of X accounts inside the Infinex mindshare cloud.

In addition to indirect attribution, we still offer referral codes. If you convince everyone in your private Telegram group to sign up, you'll receive 50% of the benefit; the other 50% goes to the individuals who dominate that day's mindshare. While seemingly counterintuitive, signups rarely derive solely from referral codes. Typically, a user's intention to sign up comes from multiple interactions and sources.

This novel campaign isn't about driving superficial attention; it aims to generate meaningful awareness for Infinex. We call this approach "Mindshare Farming."

Kaito has recently received criticism on the timeline, criticism that's partly justified, as campaigns rewarding low-quality AI-generated content harm the ecosystem. That's why we've developed a proprietary algorithm to carefully measure content quality specifically for Infinex, overlaying this quality metric onto Kaito’s raw data. While Kaito has attracted some top Twitter accounts, many have remained cautious. We believe our campaign will clearly demonstrate that high-quality content about Infinex is rewarded, and low-quality "slop" is discarded.

Of course, botting and sybil attacks are risks for any campaign. Fortunately, Kaito’s "smart follower" metric is costly to manipulate, leading to higher account and content quality. Twitter's monetisation system already demonstrates the value of high-visibility accounts; our campaign amplifies this effect tenfold. With approximately $6 million USD distributed over four seasons, I expect this to usher in a new meta and that other projects will deploy similar campaigns.

A minor note: teams are often excluded from campaigns like this. While I disqualified myself and redistributed the rewards to everyone else on the leaderboard, I felt strongly that the rest of the team should also get to participate. Having a wide variety of perspectives across the team that are actively engaged on the timeline has been a significant boost to engagement over the last year. I encourage founders to let their teams Yap.

One final point: retroactive rewards are a powerful way to prevent gaming, so that's what we opted for in Season Zero, which concluded in stealth this week. There were numerous hints and alpha drops over the last few weeks as we finalised the campaign details, but we didn’t actually freeze the scoring algorithm until the absolute last minute. If you were paying attention, you might find yourself here.

For full details of the campaign, please refer to this post.

Yapfinex