A new decentralized approach to reviewing employees

Karthik H

·

Follow

7 min read

·

Jan 30, 2023

--

Listen

Share

Introducing DERS — Decentralized Employee Review System which enables decentralized, anonymous and fair reviews.

Let’s say you are a manager trying to improve the employees' overall productivity. In terms of the general apps available, there are a ton of them from note-taking to mindfulness apps to task managers. However, let’s say that all of these things are still not working out to improve the employee’s productivity.

Then the only solution available in a large organization is to monitor employees’ and make sure they are productive enough.

Often this creates many problems as reviews can be biased and managers who do the review often won't have enough time or the will to do so.

In this Blog, we are introducing a new concept of employee review system called DERS which stands for Decentralized Employee Review System. This takes inspiration from blockchain and kleros protocol and at the end of the day helps us determine the best and the worst performing employees’ with good accuracy.

Table of Contents

1. Why DERS??
2. What is DERS?
   - Establishing some Basic terms
   - Some Important Prerequisites
   - Implementing DERS
   - Some Important Assumptions
3. One Obvious Problem
   - Enter, Kleros Protocol
   - Implementing kleros into the system
4. Demo Time
5. The inspiration behind this idea
6. Who would this Help??
7. Conclusion

Why DERS??

The current system for employee review typically involves a formal or informal process where an employee’s job performance is evaluated by their manager or supervisor. This evaluation is usually done on a regular basis, such as annually or semi-annually.

Often, such a system can face many problems such as:

1. The review process can be influenced by personal biases, such as gender, race, or age, which affects the accuracy of the evaluation and potentially leads to unfair treatment.
2. Reviews are often conducted infrequently, leading to a lack of feedback and opportunities for employees to improve and grow.
3. Reviews can be subjective, as they often rely on individual opinions and interpretations, leading to inconsistencies and potential conflicts.
4. The metrics used for evaluation may not accurately reflect an employee’s contributions or job responsibilities, leading to incorrect conclusions about their performance.

Hence, such a system usually won't work in a large organization and would lead to inaccurate results.

Our decentralized employee review system (DERS) could potentially address some of the problems that exist with the traditional employee review process leading to reduced bias, increased reviewing frequency, increased objectivity and more relevant metrics.

What is DERS?

Establishing some Basic terms

At a basic level, DERS involves the following main concepts:

1. Points and Leaderboard: The basic system involves a gamified leaderboard system which is accessible only to the admin. The leaderboard will involve all employees, with some points associated with them.
2. Cycle: A cycle in DERS ends when the results are revealed to all the employees and the good performers may be rewarded and bad performers may be punished. After this phase, another cycle starts where all employers start with 0 points.
3. Group: A group involves some randomly selected employees and such a group is formed from scratch every day.
4. Reviewee and Reviewer: A reviewee is a person in a group who is being reviewed and a reviewer is the rest of the group who are reviewing the reviewee.

Some Important Prerequisites

1. An employee would start his day by writing down a message as to what was his progress yesterday and the scrum master or the manager would verify the same. In the case of a scrum master, the verification can happen during the scrum itself.
2. Each employee is given a login ID which he uses to write down a message, which is also used by the system to uniquely identify him. This should be kept private by the employee.
3. This chat once verified is stored in the database after paraphrasing it. The reason for paraphrasing will be stated later. It also uses encryption/decryption techniques to make sure the chat doesn't get tampered with.
4. We make sure to store messages from the last few days(in our example last 15 days) as only those messages would be used for the review.

Implementing DERS

With some concepts established, we now go for establishing the concept of DERS. Consider a scenario where a company has a good number of people in each department (Marketing, Sales, HR etc) and all of them have a scrum master or manager under them.

1. A group, let's say consists of say around 20+1 employers from the same department. This group is randomly chosen every day by the system. Many such groups are made by the system on a daily basis.
2. Within a given group, 1 person is chosen by the system to be a reviewee and such a decision can be affected by the number of times that employee has been picked up by the system previously.
3. Details of all 21 people are completely anonymous and the 20 reviewers will be displayed the reviewee’s last 15 days’ messages. As the messages are paraphrased, there is no way to identify the reviewee based on writing style.
4. Now, 20 people give their reviews after thoroughly reading through the messages, after which the average score is calculated and stored in the database. If someone is already reviewed by the system, an average can be taken again to make the maximum score out of 100.
5. This happens daily and eventually, in a cycle an employee will be reviewed at least once and will have a score out of 100.
6. Once a cycle completes, the top and bottom scorers will be chosen and can be interviewed to punish or reward them.

Some Important Assumptions

1. The system makes an assumption that the admin is working for the company and will not tamper with the data or the leaderboard as doing so would only affect his own company directly or indirectly.
2. The storage of data, the process of reviewing the reviewee, and verification of the message would happen seamlessly and would be tamper-proof and for this sake, many cryptographic protocols can be used and even blockchain can be used.

One Obvious Problem

An Obvious Problem we would face with such a situation is how can we ensure that all 20 people in a group take their time and analyze the messages and then give their ratings.

Enter, Kleros Protocol

Kleros is a decentralized dispute resolution protocol built on top of the Ethereum blockchain. The Kleros protocol uses a system of smart contracts and random selection to form a panel of jurors who are incentivized to rule on the dispute. More on Kleros.

Implementing kleros into the system

1. The only change to make is to give incentives to the reviewers. Normally, Once the 20 people give their review, an average score is calculated which will be considered as the reviewee’s score.
2. What we can do is we can reward/punish the reviewers by taking the absolute difference between the score they give and the average score. The difference can be used to a very small extent to modify the reviewer’s score on the leaderboard.
3. This would give a really good incentive and purpose to the reviewer and we can ensure that a person always takes some time to go through the messages.

Demo Time

For the purpose of demonstration the following assumptions have been made:

1. We are assuming that there are 5 users in a group and all of them don't know each other. Once a user is selected, only the previous 5 chats are shown in the demo.
2. For the purposes of showing the calculation, we have added all 4 users rating interfaces on the same screen. So we supplying the inputs would be equivalent to all the users giving their ratings separately.
3. Once the calculation happens, the leaderboard is shown only for our reference. In reality, it is always hidden and would never be shown.
4. So in an actual system, the interfaces for users to give reviews would be completely isolated from each other and would in no way resemble this system. This system is for the reader to get an idea about DERS.

DEMO LINK :

DEMO VIDEO:

CODE LINK:

The inspiration behind this idea

The idea was thought of during PyHack’23 hosted by the accelerate club, RVCE. When we were thinking about the possible solutions regarding productivity we came up with solutions like

• A personalized productivity app
• A project management tool
• A time tracking and analysis tool
• A gamified employee engagement platform
• A mindfulness application

However, most of them were already developed and have been evolving. That’s when we thought of a decentralized review system which would be free from bias and after some research, we found that such solutions were still in development and a strong foundation was yet to be established. During the brainstorming session, we discovered that the incentive mechanism provided by the kleros protocol would help us out.

Who would this Help??

1. This would help any organization which is trying to achieve high productivity among the employees.
2. It would be especially useful in a large organization where anonymity can be maintained to a large extent and such a complex task can be carried out every day without any problems.
3. Decentralized employee review systems can provide employers with a more objective and accurate picture of an employee’s performance. This can help employers make better hiring and performance evaluation decisions.

Conclusion

DERS is a relatively new concept but has many advantages. It is free from bias with an increased frequency of reviews ensuring anonymity all the time.

However, it’s important to note that decentralized systems also have their own challenges, such as increased complexity in managing evaluations from multiple sources, potential conflicts among evaluators, and potential costs to implement such a complex system, especially for a small company. Additionally, decentralized systems may not address all the issues with traditional employee reviews, such as resistance to change and limited professional development.