July 5th, 2023 at 4:20:13 PM
Many recruitment teams face the challenge of hiring for a very specific profile. This may mean not just professional, but also soft skills and culture fit, and most often - a combination of all the above.
The gentle and pensive maiden has the power to tame the unicorn, fresco by Domenichino, c. 1604–05
First that comes to mind - just go on LinkedIn and reach out to everyone whose profile fits the tech requirements, and at least one of the potential candidates should check all the boxes, right?
Mass messaging wastes a lot of recruiter’s time and energy, creating a huge pool, 90% of which is irrelevant. This doesn’t leave much space to personalize the message, thus the reply rate isn’t high, which demotivates the recruiter and doesn’t demonstrate the result to the hiring manager.
In this article we will tell about our experience at Bee’s Knees of closing a one of a kind type of role.
Company: an open-source security tool for engineers
Tech profile: a senior engineer with experience in open-source and / or security products. Very strong algorithmic skills are a must
Soft skills: communication, ability to articulate their solution to others, proactive in open-source community, desire to make an impact and learn from others
Geography: anywhere in Europe, the company is remote first
We won’t lie: at first we decided to just go out there and knock on all doors, aka candidates' inboxes. We reached out to whoever had a relevant stack.
During the initial screening we would filter out candidates who didn’t have fluent English, communication skills, right motivation or had irrelevant experience. Most of the candidates submitted passed the first screening with the company and were given a test task.
That’s where the problems started. Candidates either weren’t doing the test task, or were doing it incorrectly.
Thus we identified the bottleneck and the reasons behind it:
Low motivation to proceed
Low technical skills to solve the task correctly
Taking this into account we understood that we should change the approach.
In fact, any human being is reluctant to change their approach and choose to go another way. This is natural, because acknowledging your mistakes isn’t pleasant, especially after a lot has been done, and finding a new path takes more creativity. But this ability to iterate is what differentiates good professionals from mediocre.
Even though in the beginning we reached out to some employees of the potential donors (companies that are relevant to hire from), this wasn’t our main strategy back then.
But now we decided to make it a North Star for our search. By hiring from the companies, whose culture and product matches our client’s, we could tackle both issues of low performance on the test task.
We mapped the companies according to the following criteria:
Companies that make tools for developers
Companies who hire engineers with strong algorithmic skills (insufficient knowledge of algorithms was one of the reasons previous candidates were failing the test task)
To do the mapping, we:
Looked for the similar to our client’s companies on LinkedIn (Pages people also viewed and People also follow sections)
Analyzed the companies where employees of the client worked previously
Used Crunchbase and Clutch
To find outstanding candidates who are active open-source contributors, we also used Github, including the search by the Github users lists like this one.
This resulted in the lower volumes of the messages we were sending to the candidates, also lower reply rate (because now we were targeting top companies on the market). However, the rates of passing our screening calls, company screening calls, and company’s test task, have sufficiently grown as seen below.
Before the changes:
Reply rate - 40%
Passed Bee’s Knees screen - 50%
Passed client’s screen - 85%
Passed test task - 30%
After the changes:
Reply rate - 35%
Passed Bee’s Knees screen - 75%
Passed client’s screen - 95%
Passed test task - 50%
At Bee’s Knees we are well aware that when communicating with the candidates we represent our client’s culture. This case taught us to go deeper by:
Turning screening calls into deeper conversations: building rapport with the candidate and asking more questions about motivation and previous experience
Analyzing Github profile thoroughly, asking for code examples and about pet projects
If we had any yellow (not even red or orange) flags regarding the candidate, we discontinued the process.
This made the rate of passing the company’s first screen by our candidates close to 100%.
After we made those changes, we managed to close two outstanding candidates who have already passed the probation by now. Of course, we still have a lot to learn and experiment with to be closing rare “unicorn” roles more efficiently. We actively share these experiences with the team and learn from each other, to be able to provide first class experience to our clients and candidates.