We use a combination of cultural fit, confidence, and reputation.
Cultural Fit
Cultural fit varies based on the company, team, and project so there isn’t much value I can add for you here, except that you need to determine what your values are (e.g. Quality, Value, Clarity) and find people who share them. It isn’t so important that they have the same personality, hobbies, etc., in fact, a lot of diversity is great, it is more important that they share the same values.
Confidence and the Dunning-Kruger Effect
When doing hard, new things, you have to find people who are confident in their abilities. In development, we call these abilities skills. They include languages, frameworks, paradigms, patterns, etc. We look for developers who are confident in the skills we need.
The problem with confidence is best illustrated by the Dunning-Kruger Effect. This term was coined by two Cornell psychologists who found a “cognitive bias whereby people who are incompetent at something are unable to recognize their own incompetence.” – Forbes
In one experiment, for example, they asked participants to rate how funny jokes were and then asked them how good a judge of humor they think they are. They found that those that were the worst at judging which jokes would be funny, also rated themselves the best judges of humor.
This and other studies led to the conclusion that less experienced people overestimate their abilities; in other words, they’re more confident than they should be, so their confidence doesn’t mean anything. Confidence is good, only if they have the experience to back it up.
We ask developers about their confidence in the skills we need in several different ways during pre-screening as well as in the interview, but given this human bias, we can’t just take the developers confidence at face value. We must discount their confidence rating based on their level of experience.
For instance, if our project requires a high level of confidence in React, a JavaScript front-end framework, we ask about the developers confidence in that skill in several different ways, both in ranking as well as imaginary tasks. If the developer rates themselves at an expert level, but they have only 1-2 years of experience, we must discount their level of confidence. The only exception to this is if the developer has a lot of experience in a related technology like Angular or Vue.
Reputation Economy
We’re in a reputation economy now. Gone are the days when we check references, I’m more interested in online reputation, connections, followers, recommendations, and reviews. The best guarantee that you won’t be stiffed by a bad developer is if you can leave them a review that hurts if they do.
We only end up considering about 1% of applicants. That means based on our filter questions and resume review, we only end up sending 1% of applicants to the hiring manager.
Conclusion
Filtering candidates based on fit, reputation, and confidence is the first step, testing is the next step. Learn more about how we test developers and get other tips in our free Guide to Tech Projects for Non-technical Founders.
If you need hands on help with your project, Buink has years of experience managing and completing technical projects on time, under budget, and with a high level of quality. Contact Buink today.
