How to Get Past the HR Filter For Data Scientists
The HR filter is the first layer for an aspiring data scientist to break through. Here is how to start getting more call backs and interviews.
Vin Vashishta | Originally Published: October 27th, 2017
Getting a call back for a data science or machine learning role means getting past the HR filter. I have written extensively about how to empower HR so they are able to attract data science talent. I will be honest; it does not seem like many companies are listening. From recruiters to talent acquisition, the people in the trenches hate this problem as much as candidates do.
Your job search is not going to wait for processes to catch up so what is an aspiring data scientist to do in the meantime? The more you understand the process, the better you will be able to navigate it successfully. I will start by explaining the process, so you understand what is happening when you submit your resume. Then I will explain what you can do to make it through the filter.
Step 1, the hiring manager writes a job description (JD). HR reviews the JD and marks it up with areas to fix or clarify. There are completely legitimate reasons for these markups from compliance to compensation to using language that makes it easier for the filter to do its job. A week and 5 rejections later, the hiring manager is just writing whatever HR tells them will get the job posted.
The hiring manager has 3 projects and 8 direct reports to keep track of as well as data science deliverables of their own that keep getting pushed back because they must reword a JD again. They are frustrated beyond belief. The HR person is wondering how someone in a management role cannot write a proper JD without having their hand held. They have 35 other JDs to edit for the 5th time, but they cannot let it through with a flaw or they are likely getting fired. This is not the first go round for this pair, so their fragile relationship is borderline toxic at this point.
The JD is finally approved and posted to multiple job boards. For every 2 employed data scientists, there is an open job so that JD sits on the job boards for weeks. Most applicants are under qualified, inflate their credentials, or are unscrupulous recruiters pretending to be candidates to get a contract. If you have submitted your resume, this is the stack it is in.
You wrote your resume to stand out based on the JD, but the problem is, the JD does not really reflect the job. Your resume is bounced by the filter because it does not stand out from all the others. No call back.
The filter is looking for key terms. In data science, that can be problematic because there is not really much standardization of terms. Think of loss versus cost if you want a good example. The way key term searches work is the recruiter enters a sequence of job titles and/or skills into a Boolean search query. If your resume does not have the right terms, you never show up in the results. What’s worse, spammers and scammers have figured out the most common queries. They stuff their resumes with terms, so these stuffed resumes show up highest when ranked by relevance which is usually just a keyword frequency algorithm. Your resume does not have the right or enough terms stuffed in so, you guessed it, no call back.
Month(s) have now passed. On one side, you have a team that is at least one member down so everyone has been working more to pick up the slack. The hiring manager’s boss is asking weekly about the open seat because the longer it goes unfilled, the more likely it is to be eliminated to cut costs. From an executive standpoint, the team has been getting by fine without the employee for month(s). They can save money while not impacting productivity or downsizing, win, win.
On the other side, you have a requisition that has been open far longer than average; not the sort of thing that looks good on a review. The team’s Director or VP is talking to a HR Director or VP so there’s pressure from above to get things moving.
Most aspiring data scientists are applying to jobs the day they come out. Stop that. When you see a fresh data science job opening, set a reminder for 4 weeks from now to revisit it. Why? The longer a job is open, the more enthusiastic the company gets to fill the job. As an aspiring data scientist looking for your first role, take advantage of this. No one likes pizza as much as someone who is hungry, and no one likes an aspiring data scientist as much as a company with a seat that has been open for a couple of months.
If it is still there, start looking at the existing team. Ignore the JD altogether. Remember that it has been watered down in translation. Find the hiring manager on LinkedIn. Then find a couple of members of the team with the job title you are applying for or as close to it as you can get. If you are working with a recruiter who needs to keep the company anonymous, ask them to do this for you. It is 2 minutes of work on their part and any decent recruiter will be more than willing to do it.
Build your resume up to resemble what you find in the team’s resumes or LinkedIn profiles. Use the same language and terminology. Highlight project work and education that aligns with their description of the role. If you find a lot of overlap, this is a role you have a shot at. If your resume does not have much overlapping material, do not apply.
Term stuffing is an ugly necessity. Do it badly and you will be bounced with the exaggerators. Do not do it and you will never make it into the top search results. The best way to do it is to add a keywords section to the beginning and end of your resume. Make it look like the key terms for a blog post. Most people have gotten used to seeing them, so it does not raise any red flags. It also accomplishes the goal of dropping all the relevant keywords into your resume at least twice. If you are aggressive about stuffing and twice is not enough, put terms in the footer or bottom of the page in the same color text as the background. Algorithms see it just fine. People never notice.
Another common misconception is that this is a numbers game. Applying more raises your odds, right? Applying for every job you see does is raises your frustration level, creates a lot of work for you, and possibly overexposes your resume. If a recruiter or talent acquisition specialist sees your resume for the 3rd time, what do you think is going through their head? 3rd time is the charm or hey look, Rick again.
If you do not have the right project experience yet, spend the next 4 weeks working on a sample project that aligns with what you will be doing in that role. If the job’s been open for several weeks, it will probably be there when you are done and there is never a downside to adding a practical project to your resume. Think about how it sounds to both the filter and hiring manager to say, “I didn’t have the right kind of experience a month ago, so I built this project on GitHub to prove I could be a contributor to your team.”
Be your own filter first to avoid over exposure. If the job is not a fit, do not force it. If the job is gone after 2 weeks, do not worry. Demand for data scientists is going up, not down. Do not filter on the JD, filter on the team. If you are not a fit for the team, the JD is irrelevant so do not waste your time. If this is your dream job, go the extra mile to prove you are worth at least an interview.
If you have tried everything but nothing seems to work, it is time to be realistic. You may not be ready. Spend some time thinking about what’s missing, experience or education? Ask data scientists who you are not trying to get a job with what you are missing. You will get an honest assessment of where you are and how far you still have to go. Find a Master’s, PhD., certification, or boot camp to fill in the educational gaps. Take a job in research, analytics, or software development to fill in your experience gaps.
There are a lot of aspiring data scientists who apply for their first job years too early. It is a long road into the field and your first attempt may not be successful. Mine was not. If this is really your passion, keep moving forward towards your goal.
Your first interaction with a hiring manager should not be you coming to them asking for a job. This puts you in a terrible position. Wouldn’t it be better for hiring managers to come to you? Networking is about getting people to come and talk or listen to you. That is the central concept to be mindful of when you are networking. Give people a reason to connect and they will.
Connections need constant attention to build a strong bond. That comes in the form of content. That does not mean DMs and emails with contrived reasons to talk hiding a selfish agenda. Share your projects on GitHub or by writing a post about them with visuals and code snippets. Share what you are reading and why you think it is important. If you are at a conference, share what you are seeing and hearing about.
Share with the goal of showing that you are passionate about and involved in data science. That is best accomplished by showing your growing competence in the field. Share with the target audience of hiring managers in mind. When you have got their attention, you will find that landing a job in the field is a matter of reaching out to connections to let them know you are looking. It is the ultimate way to bypass the filter. If you are successful, they will come to you with opportunities. Now you are in a position of power. You can choose who in your network you will ask for an opportunity or even which opportunities you want to explore further.
Networking is more about building a career than just landing a first job so do not neglect your network after you get hired. Continue to nurture it with content. Begin to build your own voice and prepare to ask for your next role or have your next role find you.
No matter which route you choose, networking or applying for a job, the filter is something you will have to get past. Neither route is easy so develop a short memory for failure. Learn from it and move forward. Do not be afraid to ask for what you want. However, remember that you must give people reasons to give you an interview. Also keep in mind that you are competing against a lot of other exceptionally talented people. Being successful requires more effort than mass mailing your resume.