Collabberry’s first prototype!

July 15th, 2024

Hello everyone,

We’re here to report back on our first prototype test with Collabberry and RnDAO.

Collabberry was born from the RnDAO’s fellowship program, and they’ve been our first supporters, partners, and mentors. Compensation is a very relevant problem to RnDAO as well, as they struggle with it and have done a lot of research, and designed their own system, which Collabberry learned a lot from.

It was a great milestone for us, to have our first prototype tested and potentially in future implemented by the RnDAO team.

Background

The way it all started was that in Collabberry we were working on a prototype and wireframes for the model, when a team member of RnDAO reached out, looking for some input as he was trying to solve some of the problems with their compensation model.

We quickly pinpointed some of the problems that Collabberry solves by design and decided to put the effort into building a fully functioning compensation tool, instead of a clickable prototype. We use very much centralized but out-of-the-box tools, such as Google Sheets, Google Forms, and Discord, but with that we managed to develop a fully functioning prototype.

Problem

The nature of Startups and early-stage projects is inevitably dynamic and flexible, and there’s a need for a compensation model that is as well. Negotiation and lack of transparency lead to unfairness in the team and inevitably bring tension to the team. Founders spend a tremendous amount of time doing admin related to salaries and equity packages, trying to measure value and managing the team, to make sure that everyone delivers on their promises, and dealing with legal entities for all of this to work.

Solution

Collabberry aims to bring decentralization and trust in this process, by allowing the team to measure the value brought, through peer-to-peer assessment, and introduce dynamic ownership, so that a contributor can be dynamically and on-the-go compensated with team points, rewarded for the value they brought month by month.

We aim to create a collaborative environment with aligned incentives and fair compensation.

How it works?

Collabberry’s compensation is based on an agreement between the organization and the contributor, a peer-to-peer assessment process for measuring value, and an algorithm with an outcome of a compensation package with monetary and ownership components, to reward everyone, ensuring all their essential needs are covered.

Agreement

The first input of Collabberry’s algorithms is the agreement that the contributor has with the organization. This agreement can be reached by putting up a proposal, in the DAO world, or through a standard process of communication between the contributor and founder.

The parameters of this agreement are the following:

Market Rate
Monthly Commitment
Basic Income - minimum amount of FIAT that the contributor needs to cover their costs

In Collabberry we use the model we preach, so here’s an example of one of our contributor's proposal as an example:

In this case, our berry Cansy has a minimum fiat (Basic Income) of 0 because he’s financially sustainable and happy to work only for ownership of the project.

Peer-to-Peer Assessment Process

In Collabberry’s belief, measuring value top-down is a complicated, overwhelming and also unfair process. We believe in Wisdom of Crowds and letting contributors assess each other, to measure the value brought to the team.

However, one thing to consider is that not everyone in the team has the context or the skillset, to assess others, that haven’t actively worked with them this month. That’s why we let the contributors self-state and decide who to evaluate and who not to. If they haven’t worked with someone, they won’t assess them, which practically delegates their opinion power to the ones that have the context needed.

Another thing that we considered is that in a team, the value brought is not only through getting tasks done but also it’s essential what’s the cultural contribution of each team member.

That’s why contributors assess each other based on two different aspects with 2 different scores - cultural impact and work contribution.

And last but not least, the only way for a person to grow within a team, is to learn and listen to their teammates. That’s why we introduced feedback sections in the peer-to-peer assessment process.

Here’s a screenshot to get a more practical idea of what I’m talking about.

Collabberry Algorithms

Collabberry’s algorithm is the core of our protocol. The algorithm has inputs and organization settings.

Setting: Performance Adjustment Rate (PAR)

The setting that the algorithm takes into consideration for the organization is PAR. This rate indicates how much a contributor’s salary can be adjusted based on their peer-to-peer assessment score.

The base salary of a contributor is defined by multiplying Market Rate and Commitment.

💡Example:

The PAR of an organization is set to 20%.

Alice’s agreement with the org is the following:

Market Rate: 1000$
Commitment: 100% (fulltime)
Basic Income: $500

In that case, the base salary of Alice is $1000, and the minimum compensation of Alice can be vary between $800 and $1200, and she’ll always be getting her Basic Income of $500.

If the PAR was set to 50%, then Alice’s base salary would be $500, and it would vary between $500 and $1500, and again she’ll always be getting her Basic Income of $500.

Inputs: P2P Assessment Scores

After the peer-to-peer assessment round, the algorithm takes the average of all scores received for each contributor and from there defines the Salary Adjustment Rate, based on the PAR and the contributor's score. Then it applies this rate to the Base Salary of the contributor, and 🎉🎉🎉 we have the compensation value of the contributor.

Compensation package

The final compensation package consists of FIAT and Team Points (TP), representing ownership in the project. As for contributors is essential to have their basic needs covered, they always receive their Basic Income with priority, and the extra is taken from their TP or added to their TP.

Let’s take again Alice as an example:

Market Rate: $10000
Commitment: 50%
Basic Income: $800

Let’s say that Alice has done amazingly this month, and as per her P2P Assessment scores, she’s to receive $6000, instead of her base salary of $5000.

Alice will receive her needed Basic Income of $800, and the rest in TP: $5200.

In the case where Alice underperformed, let’s say she’s to get $4500, instead of her base salary of $5000. In that case, she’d receive her Basic Income of $800 and TP: $3700.

If you want to understand better the algorithm and process, please refer to our documentation.

How did we do it?

We built the backend, calculations, algorithms, and ledger for tracking compensation on Google Sheets, and the user-facing part on a Google Form. We were back and forth with RnDAO to make sure that Collabberry in this low-tech form, brings value to the organization.

Mid June we organized the first test with the whole team. RnDAO uses Coordinape as a part of their compensation stack, the idea was to run the experiment of having another round with Collabberry, assessing the contributors for their May contributions again.

The form had a couple of feedback questions at the end to ensure we could collect relevant feedback.

Feedback

From the feedback gathered, we learned a couple of very important lessons

The process was substantially better than the compensation model they currently have.
The measurement of Quality and Quantity are radically different and should be measured separately. Basically there are two questions that need answering:
1. Have a contributor done the work committed based on their agreement?
2. What was the quality of their work and cultural contribution to the team?
That’s what lead us to the next point.
The process should give the opportunity to challenge agreements. People think about compensation during the compensation round, and then they forget about it and do their weekly tasks. It is essential to give opportunity to contributors to flag that there’s a need for adjustment in agreement when they are actively thinking about it.
There are different levels of context that you have about a contributor.
1. There are people that a contributor works with on a daily basis and have the full picture of their contributions
2. There are people that a contributor has a little context about
3. There are people that a contributor has no context about
Context is not as simple as either you have it or you don’t.
Feedback questions force people to think more about the contributor they are assessing, which makes their assessments more accurate.
UX improvements, like selecting the people you’ve worked with beforehand and then filling the form only for them rather than skipping those you haven’t worked with in the process

Next Steps

We already did a second version of the prototype, addressing all the learnings we have from the first version. We’re looking to test it with other projects and already have secured one more early-stage team.

In the meantime, we’re working with RnDAO closely on a proposal for Collabberry to replace their current compensation stack.

Reflections

We believe that we reached a crucial milestone for the development of Collabberry with this test. While we had planned to have a clickable prototype show it to projects, and run interviews, we believe that having a prototype gave us insights that we wouldn’t have managed to acquire in any other way.

After all this, now we’re working out wireframes for the platform that we’re starting to build.

Special thanks to Cansy for working out the wireframes, and to Nickolay from RnDAO who was working with us closely in the process.

Super excited about the next steps ahead of us!

Speak soon berries!

PS: In the next section, you can find all the resources and results we used for the prototype.