Recruitment in a Distributed Organisation

Recruitment in a Distributed Organisation

How to go about recruitment if you got rid of HR.

This is the story of in99, a small legaltech startup, that put in place a process to recruit high-quality professionals without a HR department and without one-on-one interviews. Doing this is part of their journey towards what they call a “Distributed Future”, which includes, amongst other things, becoming a Remote First company.

Journey towards a Distributed Organisation

Last April, in99 became of age. For ten years it co-exisited with the much older and established legal firm from which it was born. During those years the legal firm and the startup had shared resources, offices, some staff and quite crutially, central services like marketing and HR.

The coming of age of in99 was marked by the vision of being a distributed organisation, a company where teams and individuals enjoy high levels of autonomy, flexibility and responsibility, and with minimal central infrastructure. They saw their future as a team of teams, with few central functions, inspired by remote-first organisations like Automattic and Github. The 17 employees of in99 left the offices and embraced 100% remote work.

This is easier said than done. There is a lot to learn and unlearn, individually and collectively. There are many new shiny opportunities but also plenty of new challenges. At this point, they brought me in to help them design and implement a recruiting process fit for a distributed organisation.

The Challenge

One of the problems of leaving their family home is that they no longer enjoyed the benefits of an HR department, with all their skills and expertise in, amongst other things, recruitment. One initial temptation would be to find a new employee with an HR background who could take on the duties of writing job descriptions, putting up the ads, screening candidates, interviewing them, shortlisting them and organising interviews with heads of department. But, well, that would not quite fit with the vision of the Distributed Future, would it?

The key question is, how can teams do their own recruitment in a way that is effective, that attracts the best talent and that is distributed and thus scalable?

This article outlines the process we followed to address this question, and it discusses its key elements.

Key Parameters

These were the characteristics required for the new recruitment process:

1. Teams are responsible for defining the roles
2. Teams are responsible for deciding when and who to hire
3. Candidates experience a recruitment process that is innovative and fresh
4. Candidates get to experience what it is like to work at in99
5. The selection process is free(er) of bias
6. The process can scale up to thousands of candidates with little increase in workload for the teams

Overview of the Distributed Recruitment Process

From the moment a candidate applies for a job, the process can take between 4 and 8 weeks to complete. There are two parts to it. The first one is fully automated and candidate-led. The second one brings in a number of team members in a participatory workshop. There are no one-on-one interviews.

The first stage of this journey is self-led and it includes a cognitive and personality assessment (40 min), as well as a short “solo challenge” just to make sure they candidate has the basic hard skills needed for the position (60 min).

The second stage consists of a participatory workshop with 4 candidates and 4 team members, which includes a week or two of async conversation between those 8 people.

Key Elements

Highly automated pipeline

We use Workable as our Applicant Tracking System (ATS) and it costs over 3k USD per year. It allows for a certain number of automations, like email communications with the candidates as they progress through the pipeline (BIG time saver). It also comes with an API to do all the other automations that we need. Thankfully I have a developer to do that part of the work for me.

If you already use No-Code and have subscriptions to Airtable and Zapier, you could set up your pipeline with those tools and save on the cost of Workable. Having said that, we find Workable quite useful in advertising our job positions and attracting candidates. For the time being, all our candidates come from Workable ads on Linkedin.

Cognitive and Personality Assessment

Immediately after applying, candidates receive an invite to a cognitive and personality assessment. The tool we use is Bryq (3k+ USD on the first year) and it connects to the Workable automation flow. One feature that I particularly like is that we can define the personality profile that we are looking for in this particular period of time. Say, for instance, that our team of legal engineers scores very high on concrete reasoning, we can adjust the personality assessment so it gives preference to candidates who lean towards abstract reasoning, thus improving the diversity in the team. Furthermore, by doing this we (hope) we are reducing bias in our recruitment process compared to a traditional one.

The “Solo Challenge”

After the first filter from the cognitive and personality assessment, we invite applicants to complete an exercise related to the position they are applying for. It is particularly hard to create an exercise that is meaningful and that can be evaluated automatically, without any human involvement. For developers we use Codingame.eu, but we are still working out a solution for more creative positions, like UX designer. At the time of writing, we are considering Test Gorilla, which also connects with Workable.

The “Challenge with Humans”

At the end of the first stage (assessment plus solo challenge) we filter out the majority of candidates and we are left with a handful of them for the final stage, which involves current team members. Successful applicants are invited to an online collaborative workshop to work on a sandbox challenge. For now we call this step “Challenge with Humans”. It lasts two hours and there are 8 participants (4 candidates and 4 team members). The key goal is for everybody to get to know everybody else in a simulated work situation and get a feeling of what it would be like to work together. The workshop contains a number of participatory formats inspired by Liberating Structures and we make good use of Zoom’s breakout rooms. It’s important to make it fun and dynamic. The feedback from candidates has been very good every time, the find it fresh, original and that it gave them a good sense of their potential future team mates and whether they see themselves working with them.

Is it working?

Pros — It works. During the summer we added three new team members and each team was involved in the process with a reasonable amount of time invested.

Cons — Buiding the workflow with all the automations has been challenging. If you want to do this at home, you will either need a developer or someone very good at #nocode.

Some observations

• We received hundreds of applications for each position, except for the tech team. We have used paid ads to attract developers and we are still looking into how to use the fact that we are a 100% remote company to attract tech people who might be living far from the big cities.
• Over 30% of applicants don’t bother completing the Personality and Cognitive Test. On the other hand, most applicants who complete that first test also complete the Solo Challenge.
• Speaking of the Personality and Cognitive Test, we have observed some variation in the results and that some good candidates fell through the cracks. We will keep an eye on this and see how we can address this.
• Speaking of the Solo Challenge, we are still struggling to automate some of the most non-technical positions in a way that does not require a human to assess the results. This is relevant for positions such as UX designers and we are unsure if it is even possible to automate that part of the process for that type of roles.
• As we were hoping and thanks to the early-stage automations, we have not had to get involved in the recruitment process until the Async Challenge.
• In short, the pipeline processes hundreds of applicants with little human input, the final stages are fun and involve many team members, and we are finding really good candidates. On the other hand, the process does not turn around fast (4 weeks at least).

If you are curious

My colleagues and I are consultants for new ways of organising, which often includes distributed governance and autonomous teams. It’s been a blast to work with in99 to develop and implement this process and we’d be delighted to work on similar processes in other orgs. However, don’t let the word “consultant” scare you off, I’m always happy to jump a call to exchange ideas and learnings. You can reach me at manel@greaterthan.works, and you can find similar content at greaterthan.works

‍