The Synth Algorithm

It’s interesting sometimes to take a step back and think about systems as a cohesive whole. If we were to start from scratch from today, would we arrive at the same place?

Recruitment is at once a simple and difficult process. Put simply, a successful hire = getting the best available candidate into the role. But our current mechanisms are almost laughably bad at achieving this.

At present, we ask people to send us a written list (!) of their accomplishments to date and a letter (!) telling us why we should hire them. The people writing these letters have a distinct personal interest in getting the job, so they’re free to exaggerate – it’s all about getting attention. We then have to try to somehow psychically understand what sort of person each applicant is based on their writing style and their choice of font before choosing who to interview.

For years, journals, think tanks, universities and even companies themselves have agreed that hiring good people and training for skills is far more effective than hiring well qualified people who don’t fit the team or the role. Yet we persist with a system that does exactly that – the precise opposite of best practice in hiring.

This system has persisted for as long as we can remember: it’s as though we’ve given up on doing any better.


Our approach at was to start from the goal and work backwards. We did this via extensive research of best practice from industry, academia, science and world leading companies – we reverse engineered what the world’s best are doing, and systemised it for easy use.


Access to global studies, statistics and longitudinal data quickly gave us over 100,000 data points to analyse and interpret. We soon found commonalities between models and approaches that could be systemised and measured to predict candidate fit with any given position. (Note: you can read about What We Measure and why we focus on fit with role, rather than team using the link below).

Our model matches and compares candidate scores directly with each role as described by the client, providing not just a picture of who someone is – but also how well (or otherwise) they will engage with the specific circumstances of the role in question.


It’s important to note that to improve the % chance of success in hiring we must act as early as possible in the recruitment process, not at the end. In other words, if you don’t interview the right people, you can’t make the right hire.

This is a severe weakness for the current system, which tends to go through resumes very quickly (6-10 seconds per document on average), then spend a lot of time on interviews, psychometric testing, and referee reports – each of which brings its own difficulties.

Using screening as your preliminary screen at the resume stage gives you a massively improved chance of talking to your highest potential candidates, based on the best predictors of engagement, motivation, and fit.

Note: if you enjoyed this article you might also find our outline regarding What We Measure and Why interesting.

Leave a Reply

Your email address will not be published. Required fields are marked *