Where The KRE Is Headed

Written by: William Mougayar

July 5, 2022

Initial Thoughts and Direction on KRE’s Evolution

We’ve been listening to many voices across the Kin community who believe it’s time to iterate once again on the Kin Rewards Engine (KRE). The goal of the KRE has always been to be a strong catalyst for propelling the Kin Ecosystem forward towards a vibrant economy.

In the spirit of continuous improvement, we took the challenge to deeply re-examine the KRE with a new set of lenses. So, over the past two months, we have been working with an academic professional who is an expert in quantitative economics of digitization and blockchain to help us pinpoint KRE weaknesses and suggest solutions and ideas to consider. Concurrently, we also re-analyzed and revisited many previous community proposals for the KRE both recent and historic.

While this work is still ongoing, and while we are not ready to publish a new KRE proposal just yet, there is a new direction that is shaping, and we’d like to start sharing it with you, with the goal of soliciting feedback and a healthy discussion. 

First, here is a summary of the key weaknesses that were identified:

1. There is no measure of transaction “quality”. Not all use cases are created equal. The KRE does not discriminate between a quality use-case and a less effective one.

2. Observing the transaction flow of Kin is not enough. Incentivizing economic transactions via the AUB alone under-powers the economy, and creates a strong emphasis on observable (and hence gamable) metrics. For example when an app developer ‘optimizes’ for the KRE this can result in a zero-sum game to create the easiest use-case which maximizes the largest KRE reward. Furthermore, there is no unique mapping between a quality use-case and transaction flow.

3. The KRE can easily lead to a “there is no tomorrow” syndrome. Each subsequent week turns the page on the previous one, no matter what. This means for a developer, that they may not be motivated to keep improving their use-cases once they reach a certain point as the KRE is only concerned with the app’s performance over the most recent week.

4. Thresholds on transactions “value” which count towards the KRE are too low. The value of Kin micro-transactions needs to be elevated, so that more valuable use cases can ensue. For example, the KRE treats a spend of $1 worth of Kin in the app, the same as a spend of 1 Kin. 

5. App Developers are too quick to sell their KRE earnings, leaving little to get re-invested into the app economies. 

6. Overall model is too dependent on an inflationary funding trend. The Kin Foundation ends-up funding everything, no matter what happens.

Then, we considered what potential solutions and approaches could be implemented to help fix these issues. While we are cognizant of the urgency to change, we were also sensitive about limiting the magnitude of change to an acceptable minimum as a first iteration. 

We landed on 5 key areas of directional change. Here they are:

1. Tackling the quality factor via an assessment of minimum Kin integration standards that will generate an assessment score. An app which ticks all the boxes with their use-case and contains all the expected integration standards will boost their earning potential from the reward pool.  Apps which are missing all or some of the integration standards, or have generated a lower assessment score, may find themselves at a disadvantage.

2. Raise all Earn-Spend thresholds. This would be a bold move, but a necessary one.  We are intending to introduce thresholds on transactions as to whether they are counted for KRE purposes. We are looking at requiring spends to be at least 2.5 cents (later to 5 cents) worth of Kin to be a valid spend transaction, the minimum AUB per user to 3x the wallet creation cost, and increasing the maximum AUB cap per user to $10 worth of Kin.  How this would work is the Kin Foundation would publish at semi-frequent intervals the minimum thresholds in whole amounts of Kin. The result is developers will be free to implement whatever Kin use-case they like, but for the purposes of calculating the KRE, only economically significant transactions above the threshold requirements will be counted.

3. Burn a % of Kin as a factor of KRE rewards. To begin with, we are looking at burning the weekly volatility factor. As you would already know, KRE payments are adjusted to account for volatility in markets. This Kin is returned to the Foundation’s reserves and forms part of its annual budget, therefore it often finds its way back out into the circulating supply by other means when redistributed through other programs.  Distributing this Kin into the circulating supply then burning it would result in approximately ~12% of all new Kin being burnt, permanently reducing the total supply.

4. An app’s ranking in relation to others for its share of the reward pool will be based on its Economic Contribution Score (ECS). This figure is calculated using all of the relevant factors the KRE will consider, such as the app’s Kin economy and performance (spends, AUB etc), its assessment score and other existing adjustment factors like the anti-monopoly provisions. Ranking apps by their overall ECS will showcase which apps were the biggest contributors to the Kin economy for the relevant period, rather than how well they performed on a single metric like AUB. For the avoidance of doubt however, AUB (as modified by the above) will still be an important metric in determining an app’s ECS, amongst the other factors. 

5. New app boost. The KRE algorithm will give a time-limited boost to apps which are new to the Kin Ecosystem. This boost will enable newcomers to kick-start their Kin economy and build momentum behind their use-cases to compete against established players. 

There you have it. We intend to publish a formal proposal soon that will cover the above in detail. There are other minor changes contemplated at this time not covered in this blog post. Therefore please treat this as high-level and directional mostly for discussion purposes. 

We’d love to hear from the community. We’ll be following Reddit, Telegram and Twitter for your feedback and discussions.


About Kin

Kin is a decentralized cryptocurrency purposely designed to integrate easily across mobile and web apps, with a built-in incentive model that rewards developers for increased usage. Apps built with Kin get paid for creating compelling cryptocurrency-based user experiences, where greater engagement results in shared economic benefits for users and developers. Today, the Kin Ecosystem boasts 65 million wallets, and has distributed over $70M in rewards across 80+ apps since its inception in 2017. Kin is an SPL token on the Solana blockchain, enabling consumer-scale apps to transact swiftly, with minimal-to-no fees. Learn more at Kin.org.

Build with Kin 

Apply for a Grant (Kin Catalyst Fund) 

Sign up for the Kin Monthly Newsletter

Sign up for the Kin Rewards Engine (KRE) Weekly Payout Snapshot

Follow Kin Ecosystem on Twitter / LinkedIn / Facebook / Instagram / Discord / Reddit