In assessing this selection process, my team and I reviewed more than 100 jobs from this year’s first quarter and analyzed the 4,242 applications received for those positions. Typically, our jobs receive an average of 14 applications but for the purpose of our study, we selected jobs that received at least 15. We needed a large sample size to better understand and identify differences in behavior, and to ensure that the results were statistically significant. So, what did we learn?
- 17% of the applications received a “thumbs-up” rating, meaning they were a fit for the job
- 39% received a “thumbs-down” rating and, therefore, were not a fit for the job
- 44% of the applications were not rated at all
We studied 20 different variables for each job application in order to identify what caused an application to be rated a “thumbs-up” (a fit) or a “thumbs-down” (not a fit), which generated a fair amount of noise and inconclusiveness.
Ultimately, we were able to identify that the variables that best explain the ratings were:
- Time to apply
- Job Title
Shouldn't Salary be a good indicator of “Fitness”?
Below are three different box plots that visually represent the salary distribution of the applicants to three different jobs. In two cases, we can see meaningful differences between “thumbs-up” and “thumbs-down,” but little difference with the non-rated candidates. In the third case, we observe hardly any difference between all three groups.
Across all 100 jobs, we did not see any evidence other than if an applicant’s salary is two standard deviations away from the average salary of the applicants rated a “thumbs-up” (a fit) than in those instances salary was a factor. In others words, for the outliers, salary was a factor that influenced the rating negatively (either low salary or high salary).
For example, with a marketing manager job where the average applicant earns $120K a year, if an SVP in marketing earning $220K were to apply to that job then, the high salary of that person would be the reason why this applicant would be rated not a fit.
The other reason why our data excludes compensation from being a negative factor is that we already control that factor in our user experience. At TheLadders, we have organized our candidates and jobs in five salary bands. The distribution is as follows:
Therefore, a professional earning $60K a year cannot apply to a job paying $150K and vice versa, a $200K Vice President cannot apply to a $75K associate job.
Example: Managing Consultant, Chicago IL
The location of the applicant impacts the candidate’s fitness for a job. This factor is specific to the preference of the employers performing the search. In some cases, the company is willing to relocate out-of-state candidates and, in other cases, the company is not. The maps below illustrate the case of a recruiter not wanting to relocate candidates. Applicants outside of Illinois were rated a no fit or simply not rated at all.
Example: IT Infrastructure Program Manager Windows 7
The data reveals that with most of the jobs, the applicant’s current title is a great variable to predict job fitness. In our example below, you can see that the applicants with titles such as CIO, VP of IT Services, Director IT, and Personnel Manager were rated not a fit or not rated at all. The recruiter behind this job is rejecting the over-qualified candidates who are applying to position below their levels (see graph below).
Example: Project Execution Consultant in Construction/Real Estate
Function was a less obvious variable that helps predict a candidate’s job fitness because we already control for function at TheLadders with our matching algorithms. However, with keyword search, we do not prevent candidates with a background in a particular function to view and apply to jobs in a different function. As you can see in the example below, the recruiters rejected candidates with a background in operations, general management or engineering. Only candidates with a background in real estate and construction received a “thumbs-up” from the employer.
Time to Apply:
Example: Division Finance Manager, Atlanta, GA
As we investigated 20 different variables such as function, salary, location of the candidate, and job titles, we found that one of the leading success indicators to be rated a fit was time to apply. What does time to apply mean? It is the number of days between the time the job was published on our site and the time (stamp) of the application.
On average we saw the following:
- 10.38 days for a “thumbs-up”
- 13.53 days for a “thumbs-down”
- 19.79 days for the not rated
The causality between that variable and the three other rating types became even more apparent when we looked into the median time (midpoint of the distribution):
- 2.81 days for a “thumbs-up”
- 4.32 days for a “thumbs-down”
- 13.95 days for the not rated
In other words, half of the applicants rated a fit for the job had applied to that same job within 72 hours of the jobs being published on our site. The median time for candidates rated “not a fit” was 50% higher and for the non-rated candidates, the median time was 400% higher. Below is the applicant distribution by time for a Division Finance Manager job in Altanta. The green bars represent the applicants rated “thumbs-up.”
Who is getting the job in the end?
The findings from this study teach us that the people who are getting the job cannot be explained by only one variable. It is usually a combination (one or many) of having the appropriate salary, the experience level, being from the appropriate function (marketing professional applying to a marketing job), having the required expertise (online marketing vs. public relations) and preferably being located near the job location (to avoid occurring relocation expenses).
However, the most eye-opening finding in our research was the following: regardless if one might be the perfect fit for a job, the later one applies to a job, then the less likely one is to get a call-back. That is the sad reality, even if that person was the purple squirrel that the company had been looking for all along.