PM Quest · Practice cases · Learning resources

LinkedIn Job Applications Spike, but Recruiter Engagement Drops

Root Cause Analysis · advanced

LinkedIn's latest A/B test shows that job applications through the platform increased by 25%, but recruiter engagement with applicants decreased by 10% over the same period. This unexpected result is puzzling, as higher application volumes typically lead to more recruiter activity. You have one data analyst available for a week. Diagnose the root cause of this anomaly and determine which data points to prioritize for investigation.

Write your approach on PM Quest

Community approaches

Pratyusha’s approach

clarifying qsns

  1. what exactly was the A/B experiment? i will start by understanding what change was made on the platform to analyse how it might have affected user behaviours
  2. how exactly are we defining recruiter engagement here?
  3. over what time frame did the change in these metrics happen?

assumed answers: What exactly was the A/B experiment? A new "Easy Apply" flow was launched that reduced the number of steps required to submit an application. How are we defining recruiter engagement? Percentage of applications that receive at least one recruiter action (profile view, message, shortlist, interview request, etc.) within 14 days. Over what time frame did the change happen? The A/B test ran for 4 weeks. The +25% applications and -10% recruiter engagement were measured over that period.

i will first start by understanding if this double metric movement is a coincidence or a causation

  1. causation : if applications increased & recruiter actions decreased only for jobs were easy apply has been enabled. else it is most likely a correlation and i would analyse it differrently

assuming its a correlation

i will now analyse user journeys to understand what could lead to change in both metrics applications : open linkedin > search / find a job > read JD (optional) > click apply > complete application > submit (target metric) > repeat user journey (optional)

  1. hypothesis that would impact total # of applications on platform 1.a) higher # of jobs or higher # of active candidates on platform 1.b) increase in ease of applying (most likely root cause) 1.c) seasonal fluctutaions in job application behaviours. for ex- end of financial yrs, layoff periods etc recruiter engagement : post a job > add application link > get applications > review applications (target metric) > continue user journey
  2. hypothesis
  3. a) increase in candidate influx- leading to recruiter exhuastion. metric to check: increase in applications per job & decrease in recruiter actions in the same job post. estimated increase in time taken to review total application
  4. b) decrease in candidate quality : no of applications with relevant skills & experience have reduced significantly post new feature release leading to recruiter disengagement with platform. metric to check : change in % of applications with high match score per job

i would prioritise checking the correlation vs causation hypothesis first and then check 1.b and 2.a and then 2.c to get to the root cause of this issue