3 ms·

At my workplace (of 10+ years), when we recruit for entry-level positions, we typically get 15-50 applicants. We are not a software company so we do not use de

At my workplace (of 10+ years), when we recruit for entry-level positions, we typically get 15-50 applicants. We are not a software company so we do not use deep algorithm testing as part of our evaluation process. Most applicants (probably 80% or more) have a computer science or information systems 4 year BS degree. The other 20% typically have a STEM degree with some programming courses or experience. Probably 20%-30% have some type of graduate degree as well - we sometimes have career changers who have picked up a MS in computer science from some coursework-only MS program at a small school. My experience has been that of that candidate pool, we will have 3-6 candidates with well documented programming projects or job experience. These tend to be the candidates we interview - we use a simple rubric with yes/no for 5-10 categories we evaluate. The categories are broad ("Has this candidate written code in either personal projects or previous employment?") and minimally skill-specific. We do tend to have some bias towards database (SQL) skills and some hands-on experience with Linux, but not specific "hard" requirements for either (don't care if MS SQL Server, Oracle, Postgres, or Debian vs Ubuntu vs CentOS, etc). We have had success with non-traditional candidates - we have team members with backgrounds ranging from BS in History to PhD in Bioinformatics, so we aren't super-credential sensitive. All these details seem to add up to a rather stark observation - there is NO shortage of highly qualified applicants for these types of jobs. Our geographic region has many tech employers so we aren't the only option for job seekers. We probably aren't even in the top 25% of employers for salary in our area. Candidates with only three months of training are just not going to be successful getting to the front of the hiring line in the deeper pools of applicants we get.

908B64B197 6y ago
> All these details seem to add up to a rather stark observation - there is NO shortage of highly qualified applicants for these types of jobs. Our geographic region has many tech employers so we aren't the only option for job seekers. We probably aren't even in the top 25% of employers for salary in our area. Where is that? > We are not a software company so we do not use deep algorithm testing as part of our evaluation process [...] My experience has been that of that candidate pool, we will have 3-6 candidates with well documented programming projects or job experience. These tend to be the candidates we interview - we use a simple rubric with yes/no for 5-10 categories we evaluate You should try it. Right now it seems you are skewing toward hiring folks who already have experience vs fresh grads for entry level positions. Whiteboard, if done right, gives a chance to college hires to prove themselves. Especially if there are other tech companies in the area. The pool of candidates that have one year of experience and are already looking to jump ship is very different than fresh grads (where everyone is looking for a job).
- clusterhacks 6y ago
  Middle-eastern United States area. Our rubric is for deciding who to interview. We definitely have new grads who have resumes with meaningful programming projects who get into the interview pool. We have hired new grads over experienced candidates on the strength of the interview. During an interview, we usually have a whiteboard abstract design and pseudo-code guided exercise. I also very much like to ask candidates to walk us through one of their projects and experience with that as well. Occasionally, someone who doesn't do very well with other parts of the interview will really open up on a school or personal project and show unexpected insight and ability. I wish we could do a blinded work-sample test, but there are HR policies that prevent us from using one.
  - 908B64B197 6y ago
    > During an interview, we usually have a whiteboard abstract design and pseudo-code guided exercise. I also very much like to ask candidates to walk us through one of their projects and experience with that as well. Occasionally, someone who doesn't do very well with other parts of the interview will really open up on a school or personal project and show unexpected insight and ability. That's the correct way of doing algorithmic screening. Relatively self-contained problem that has several possible solutions.

908B64B197 6y ago

> All these details seem to add up to a rather stark observation - there is NO shortage of highly qualified applicants for these types of jobs. Our geographic region has many tech employers so we aren't the only option for job seekers. We probably aren't even in the top 25% of employers for salary in our area. Where is that? > We are not a software company so we do not use deep algorithm testing as part of our evaluation process [...] My experience has been that of that candidate pool, we will have 3-6 candidates with well documented programming projects or job experience. These tend to be the candidates we interview - we use a simple rubric with yes/no for 5-10 categories we evaluate You should try it. Right now it seems you are skewing toward hiring folks who already have experience vs fresh grads for entry level positions. Whiteboard, if done right, gives a chance to college hires to prove themselves. Especially if there are other tech companies in the area. The pool of candidates that have one year of experience and are already looking to jump ship is very different than fresh grads (where everyone is looking for a job).

clusterhacks 6y ago
Middle-eastern United States area. Our rubric is for deciding who to interview. We definitely have new grads who have resumes with meaningful programming projects who get into the interview pool. We have hired new grads over experienced candidates on the strength of the interview. During an interview, we usually have a whiteboard abstract design and pseudo-code guided exercise. I also very much like to ask candidates to walk us through one of their projects and experience with that as well. Occasionally, someone who doesn't do very well with other parts of the interview will really open up on a school or personal project and show unexpected insight and ability. I wish we could do a blinded work-sample test, but there are HR policies that prevent us from using one.
- 908B64B197 6y ago
  > During an interview, we usually have a whiteboard abstract design and pseudo-code guided exercise. I also very much like to ask candidates to walk us through one of their projects and experience with that as well. Occasionally, someone who doesn't do very well with other parts of the interview will really open up on a school or personal project and show unexpected insight and ability. That's the correct way of doing algorithmic screening. Relatively self-contained problem that has several possible solutions.