For this post, I’ll use a framework that I created for BeyondScale‘s tech recruiting. I customized an example from Dan Kahneman’s book, Thinking, Fast and Slow with a few of my inputs.
Step 1: Define your evaluation criteria and weights
BeyondScale team felt that raw technical skill is 30% of the signal for a great employee. Starting with that insight, I added common attributes that are expected from a developer and assigned the weights.
The good thing about this framework of thinking is that you don’t have to be perfect on Day-1. If it’s not working, your team has a direct feedback for necessary changes.
Criterion | Description | Weight |
---|---|---|
Raw Technical Skill | React/Javascript/Python/Citrix/other raw skills that they use day to day in this role | 1 |
Engineering decision making ability | e.g. Evaluating and choosing libraries | 0.7 |
Git Workflow Familiarity | Important for devs to collaborate with other devs | 0.7 |
Predicting failure scenarios | Defensive programming techniques | 0.7 |
Clarity of Communication | Vocabulary, Precision and to a lesser extent Pronunciation | 0.7 |
Common sense | Not believing bullshit just because the manager said so | 0.7 |
Honesty | CVs are inflated. How much is real? | 0.7 |
Passion for great work | Great craftsmen are fundamentally driven by a need to do great work | 0.7 |
Curiosity | Tinkering with things | 0.7 |
How far have they come? | If the candidate is from a disadvantaged circumstance, then how far they came is a great signal of how far they’ll go. https://medium.com/incerto/surgeons-should-notlook-like-surgeons-23b0e2cf6d52 | 0.7 |
Step 2: Create a google form for these attributes
For this example the form looked like this – https://forms.gle/HXo8aZ96TkPxw3QRA
Post every interview, the interviewers had to fill in this form independently and restrict their evaluation in each question to the specific question being asked.
Step 3: Compute a final score for each candidate
The formula for the final score is
SUM(SCORE_UNDER_CRITERION x CRITERION_WEIGHT)
Step 4: Sort by final scores and make an offer
The final step is to just make an offer to the highest scorer. That turns out to be surprisingly hard for teams as all kinds of emotional reasons come in at this stage.
Since recruiting is a highly subjective matter, you should resist the temptation to override the final scores for emotional reasons.
Measure and tune
As time progresses, adjust the criteria and weights in Step 2 as you gain better understanding of your talent needs. This framework of thinking gives you a way to
- tap into your team’s collective intuition
- measure effectiveness of your recruiting methods
- improve your talent spotting ability by removing bias
Let me know your thoughts
Hope that helped. Feel free to share your thoughts & concerns about this framework. Do you have such criteria and weights that you can share in comments?
I’m happy to help create this framework for your team. Contact me at manu AT bountyhiring.com