Companies hiring analytic talent have been complaining for a few years about how hard it is to find good data scientist. Our analysis of job board data (based on Indeed, Datashaping and Analyticbridge statistics) reveals that a job ad specifying two possible locations - e.g. San Francisco, CA or Boston, MA - generates on average 5 times more page views, than a job ad targeting a single location.
The impact is even stronger on Indeed.com, as job ads specifying multiple (say two) locations show up to all users in US, not just those from the two specified locations. Getting candidates to see your job ad is the very first step to attract talent, and thus this "multiple locations" feature is by far the most potent solution to boost success - more potent than the job title itself or requirements.
Of course this assumes that your company can accommodate employees in multiple locations. It also works best if these locations include the Bay Area (don't be specific by saying San Jose, San Francisco or Walnut Creek) and "Greater New York" (don't be specific by saying NYC or New Jersey or Long Island). The contrast is even more striking if your headquarters are located in (say) Benton, Arkansas (e.g. Walmart). In this latter case, you would have to relocate a candidate from outside Arkansas and pay salaries higher than in New York City to attract a decent data scientist.
Besides the "multiple location" factor which is now leveraged even by the smallest start-ups including Analyticbridge, one important source of "candidate leakage" or bottlenecks is black holes in the online screening process. We will give two examples: