It seems like there are more candidates than ever before applying for open data science roles. Unfortunately, many—if not most—of those candidates are simply not qualified. So, how do you screen through all of those data science resumes quickly and effectively? We’ve collected some of our best tips to cut through the noise of an oversaturated applicant pool to find the strongest data science candidates in today’s market.
5 Tips for Screening Data Science Resumes
While today’s ATSs are equipped with advanced filters for screening and scoring resumes, we find that the complexities of data science, analytics, and engineering roles (not to mention the keyword stuffing that’s rampant these days) necessitate that an experienced data science recruiter reviews each resume as well. Our tips are designed to help you maximize your recruiters’ time and efficiently wade through the volume of resumes you receive.
1. Leverage your ATS.
You’ll likely have some filters in place before you actually have human eyes look over resumes. If your ATS can automatically reject candidates based on certain criteria you give it when configuring a new role, go for it. It’s helpful to screen out candidates who don’t meet the bare minimum requirements for the role at this stage. If we can save our recruiters time by screening out these candidates, we do!
2. Keep an eye out for red flags.
Say your ATS isn’t able to automatically reject candidates based on certain criteria you give it, or that criteria are more nuanced than an AI might be able to handle. When any of these red flags occur, we typically reject the candidate immediately and focus on the resumes that appear more legitimate:
- Missing important details: Applications where the candidate does not list their last name, or has shortened their last name to a single letter are generally red flags we look out for.
- Has not completely filled out the application: We frequently see applications where candidates will either leave our questionnaire blank, or include single letter answers instead of a real response.
- They have written responses, but something’s not right: We have been seeing a rise in AI-generated responses to some of our questionnaires. While this isn’t necessarily an automatic red flag, if the answer they have given doesn’t make sense contextually, or is overly stilted and making you scratch your head, it’s probably better to reject that one.
- Their name looks familiar: Candidates who have applied to the same role multiple times do stand out to our recruiters, but not necessarily for the right reasons. This is another flag that isn’t automatically worthy of a rejection, but it does make a candidate appear less legitimate, especially if their application has other red flags in it.
3. Proceed with caution if you see keyword loading.
Keyword loading, or keyword stuffing, refers to the practice of inserting specific words that an ATS is likely to filter for as many times as possible in a resume. While often more of a beige flag, since it’s so common in the field, keyword loading can detract from the veracity of a candidate’s credentials. We typically reject applications that are heavily keyword-loaded if they also do the following:
- Lose the thread: sometimes we come across resumes that have no logical thread between two bullet points describing a role. If the logic doesn’t wash, we reject it.
- Take up too much space: another common symptom of keyword loading is candidates cramming too much information in under the details for a role. We typically reject a resume if a candidate has more than eight bullet points per role. If they can’t cogently describe what they do in under eight bullet points, that’s a bad sign!
- Lists multiple technologies for one function: Obviously, Data Scientists use multiple technologies frequently, and if the tech stack listed makes cohesive sense, that’s great. We are curious, however, when someone lists that they’ve worked with GCP, Azure, and AWS cloud infrastructure all in the same role! Yes, it happens, and if a cloud migration is mentioned, or something similar, that makes sense. However, if those three technologies are listed for the same role with no context or explanation, that’s definitely something we would ask the candidate to explain.
4. Look for green flags in a resume.
Even with all the misrepresentation that’s common in resumes, it’s important not to focus solely on the deal-breakers. Looking for these green flags can help you identify the most promising resumes in the pool quickly.
- Longer role durations: We really like to see candidates who have held roles for longer than a year. Typically, if we see a resume with some promising skills and experience, but the roles are all less than a year old, we will reject them. Of course, this might be a bit different if the role in question is contract, but it’s also a great sign if we see candidates who have held contract roles for longer than a year.
- Relevant industry experience: This might sound a bit obvious, but candidates with relevant industry experience stand out in a large candidate pool. The clearer that fact is communicated on a resume, the more likely we are to look deeper into that candidate’s profile.
- Great answers: One of the red flags mentioned previously is when someone either doesn’t answer our questionnaire, or has relied too heavily on ChatGPT or another LLM. We love to see organic and original responses to our questions! It’s unfortunately less common than we’d like, and it’s an automatic green flag from us!
- Clear, well-organized, and concise resumes: In the age where keyword-loading is the norm, it’s frankly always pleasant to look at a more paired-down resume. If an applicant is able to articulate their experience without all the other noise, that’s a green flag.
5. When in doubt, Google it.
Does a line seem out of place in a resume? Maybe it’s written with a different tone than the rest or does not quite make sense with the other bullets. It’s surprising how often resumes include lines “borrowed” from other resumes and job descriptions on the internet. When we see a line that seems off, we frequently turn to Google. Just be sure to use quotes around the entire line in the search field, then see what comes up. I will note that this can be more of a beige flag if real competency is evident in other elements of the resume or if the line is pretty generic in its wording
Remember, screening data science resumes is more art than science.
We’ve shared our top tips to filter through resumes in a candidate-heavy market, but we also want to leave you with a few words of encouragement. It’s easy to feel a bit disillusioned when screening resumes for roles in the data sciences and other hot industries. Don’t forget that sometimes, people with very specific skills simply aren’t great at writing resumes. So use your discretion. If a resume throws up a couple of red flags, but you still see promise for other reasons, be it enthusiasm, longevity, location, or even the fact they bothered to write a cover letter, it may be worth going with your gut. Happy recruiting!