Metrics for Google's Employee Referral Channel: Hire Quality, Cost and Incrementality

You are interviewing for a product manager role at Google. The interviewer asks how you would tell whether the company's employee referral program is effective. Through the program, current employees recommend people they know for open roles. Explain what effectiveness should mean for this program, which metrics you would track, and how you would use them to decide whether the program is working. ```hint Follow the referral past the submit button A submitted referral is only the start. Think about what happens to referred candidates at every later stage, what the program costs, and what hiring would have looked like without it. ``` ### Clarifying Questions - What is the program mainly meant to achieve: more applicants, stronger hires, faster or cheaper hiring, or reaching people who would not apply on their own? - Does the program pay referrers a bonus, and is it paid at hire or only after the new hire stays for a set time? - Should the measurement cover every role and country, or one job family such as engineering or sales? - Does the applicant-tracking system record referral as a source, including when the referred person had already applied independently? ### What a Strong Answer Covers - A definition of effectiveness tied to the program's hiring goal, leading to one primary success metric. - The full path from employee participation through referral, interviews, offer and acceptance to the new hire's performance and retention. - A like-for-like comparison with other hiring channels and a way to estimate the program's incremental effect. - Guardrails such as diversity of the candidate pool, a consistent hiring bar, recruiter workload and cost per hire. - Handling of outcomes that only mature months after a hire, and of segments by role, level and region. ### Follow-up Questions - After the referral bonus is raised, referrals double but the share of referred candidates passing interviews drops. How do you decide whether the change was worth it? - Referred hires stay at the company longer than other hires. Why might that not show that the program causes better retention? - How would you design an experiment to learn whether the program adds hires or only shifts them from other channels?