I've seen engineering hiring treated as both a process and a numbers game. Without measuring the funnel, managers are left making decisions based on feeling rather than data.
Measuring the hiring funnel is crucial. We track metrics from application to offer, including application-to-phone-screen conversion rate, phone-screen-to-onsite conversion rate, onsite-to-offer conversion rate, offer acceptance rate, and time-to-hire. By tracking these metrics by role type, hiring manager, and recruiting source, we can identify where the funnel is leaking.
A low offer acceptance rate often indicates a compensation or process problem, while a low onsite-to-offer rate can indicate interviewer calibration issues. These metrics help us pinpoint the problems and make data-driven decisions to improve the hiring process.
For example, we once noticed that our onsite-to-offer conversion rate was significantly lower for candidates sourced from LinkedIn compared to those from employee referrals. Upon further analysis, we found that the interviewers were not calibrated to assess the skills of candidates from non-traditional backgrounds. We adjusted the interview format and provided training to the interviewers, which resulted in a 20% increase in onsite-to-offer conversion rate for LinkedIn-sourced candidates.
Sourcing diversity is another critical aspect of hiring. Passive candidate sourcing, such as posting on LinkedIn and waiting, produces a narrower and more homogeneous candidate pool than active sourcing. By measuring sourcing by channel, including inbound, LinkedIn outreach, employee referrals, university recruiting, diversity job boards, and the conversion rate by channel, we can identify which channels are effective for which role types.
Employee referrals typically have high conversion rates but low diversity, while diversity-focused sourcing requires active outreach. This data helps us adjust our sourcing strategies to attract a more diverse pool of candidates. For instance, we increased our outreach efforts to diversity job boards and saw a 30% increase in diverse candidates in our pipeline.
The predictive validity of the interview process is also measurable. We track performance review scores of hires 6 and 12 months after hire and correlate with interview feedback. This helps us identify interviewers whose positive feedback predicts high performer hires, and those whose feedback does not correlate.
Using this data to calibrate interviewers, adjust interview formats, and retire unproductive interview stages improves hiring quality over time. It's a continuous process that requires ongoing evaluation and refinement. We use tools like Google Forms and Airtable to collect and analyze interview feedback, and Tableau to visualize the data and track trends.
Every candidate who interviews at our company forms an opinion of it, and candidates who have a poor experience are likely to share that experience publicly. The engineering talent market is small and interconnected, so the investment in a respectful, well-organised, prompt hiring process produces compounding returns in candidate quality and employer reputation.
The cost of a poor hiring process is not just limited to the current candidate pool. It can also damage our reputation and make it harder to attract top talent in the future. By prioritising candidate experience and investing in a well-organised hiring process, we can build a strong employer brand that attracts the best engineers.