You’ve got a list of open jobs you’re interested in; now it’s time to let the employers know you exist! Pretty much every job will require you to submit a résumé: a glorified list of your skills and experience. Most jobs also ask for a cover letter: a one-page letter describing why you should be considered for the job. It would be easy to jot down your previous jobs quickly and write a boilerplate letter saying that you’re interested in the company, but in this situation, putting in more effort can be the deciding factor in whether you make it to an interview.
In this chapter, we start with making sure that your base résumé and cover letter are as effective as possible, covering best practices and common mistakes to avoid. Then we show you how to take that “master” résumé and cover letter and refine them for each job. Finally, we show you how networking can help get your carefully crafted application into the hands of a hiring manager instead of into an overflowing pile of résumés.
The only goal of a résumé is to convince a person who’s barely skimming your résumé that you are worth interviewing.
The key theme throughout this chapter is that you need to convince a person quickly that you’re qualified for the position. Company recruiters often get hundreds of résumés for each data science opening. Furthermore, because data science encompasses so many distinct types of jobs, the range of skills of people who apply for the positions will be huge. This act reinforces the notion that your materials have to say “Hey, you reading this, you can stop skimming this massive pile, because you’ve found the person with the skills you’re looking for.” But being able to show that you’re qualified isn’t an easy task.
Although networking and personalizing your application take time, they yield much better results than spending just an hour writing a basic cover letter and résumé and then applying for dozens of jobs with one click. You’ll be more likely to get an interview, as you’ll have matched your application to the company’s requirements. And when you reach the interview (the topic of chapter 7), you’ll be able to give a great answer to the common question “Why are you interested in this role?”
The goal of your résumé isn’t to get you the job; it’s to get you an interview. Recruiters who run the interview process get in trouble if they bring in people who clearly don’t meet the qualifications for the job, and they’re praised when the people fit the qualifications well. Your résumé needs to show the reader that you meet the requirements for the position so that the recruiter is comfortable moving you on in the process.
That goal is very different from creating a catalog of every experience you had, which unfortunately is a goal of many inexperienced résumé writers. Although you do want to avoid having gaps on your résumé by leaving off recent jobs, you can spend less time on those that aren’t related to data science. And even if you have a lot of data science experience, you should still focus on highlighting the most relevant jobs. If you have a multipage résumé, most recruiters won’t have time to read all of it; neither will they be able to tell which parts to read. No one will tell you “Well, we would have hired you, but you didn’t put your lifeguarding job in high school on your résumé, so we couldn’t.”
There’ll be plenty of time later in the interview process to go through your jobs, education, and data science projects in depth. For now, you want to focus on what’s most relevant for meeting the qualifications of the position you’re applying for. During the rest of the process, you’ll focus on your great qualities that will help you stand out from the other applicants, but for the first step, it’s good to focus on fitting into the hiring manager or recruiter’s expectations.
With that in mind, we’ll walk through the basic structure of a résumé, how to create good content within that structure, and how to think about the many résumé rules that get thrown around. Plenty of this content could apply to any technical position in industry, but we’ll focus as much as possible on what is unique about data science. We’ve also made an example résumé for you to use as a guide (figure 6.1).
In this section, we walk through each section of the example résumé, going into more detail about what you want to include.
Including your contact information is necessary so that the recruiter can contact you! You need to include your first and last name, phone number, and email at a minimum. Beyond that, you can put links to places where they can find more information about you, including social media profiles such as LinkedIn, online code bases such as GitHub, and personal websites and blogs. To figure out what to add, ask yourself this question: “If someone clicks it, would they think more highly of me?” A link to your project portfolio from chapter 4, for example, is a fantastic thing to include. But a link to a GitHub profile that’s empty save for a clone of a tutorial project is not. If you have any data science work that’s publicly available, try to figure out a way to show it here.
Generally, you also want to include the city and state you live in, which will let the recruiter know that you’re nearby and can commute to the job or that you’d need to relocate if you got the job. Some companies are hesitant to relocate new hires because of the expense, so if you don’t live nearby and don’t want to bite that bullet, you could potentially leave your location off.
If your legal name doesn’t match the name you commonly go by, you can use your common name. Farther along in the process, you’ll need to let the company know what your legal name is for things such as background checks, but you aren’t required to use your legal name when applying.
Finally, don’t use an email address that’s potentially offensive (i_hate_python @gmail.com, for example) or something that might expire (such as a school email address).
This section is where you show that you’re qualified for the job through previous jobs, internships, or bootcamps you’ve done. If your past jobs are related to data science, such as software engineering, that’s great: spend a fair amount of your résumé on them. If a job isn’t related to data science, such as being an art history professor, you should still list the jobs, but don’t spend much time on them. For each position you’ve held, list the company name, the month and year of your start and end, your job title, and at least one bullet point (two or three for the most relevant jobs) describing what you did. If you’re a new or recent graduate, you can include internships and research jobs in college.
This section should be the largest in your résumé and could potentially take up half the space available. It’s also often the most important, because it’s the first place recruiters will look to see whether you have data science experience that could be related to the job they’re hiring for. Due to the importance of getting this section right, we’ll go in depth into how to create the best content for it in section 6.1.2.
In this section, you list your educational experiences, ideally to show that you have a set of skills that would be useful for the data science job. If you went to school past high school, even if you didn’t get a degree, list your school(s), dates (same format as your work experience), and your area of study. If this job will be your first one out of school, and your grade-point average is high (above 3.3) you can list it; otherwise, leave it off. If you’re a recent graduate, and you took statistics, mathematics, or computer science, or any other classes that involved them (such as social science research methods or engineering), you can list those classes.
Recruiters will be very interested to see whether you have an area of study that’s relevant to data science, such as a degree in data science, statistics, computer science, or math. They’ll also be interested to see the level of your degree. Because many data science topics aren’t covered until graduate levels of programs, having a graduate degree will help. Recruiters generally won’t care about what school you went to unless it’s extremely famous or prestigious, and even then, this credential won’t matter if you’re more than a few years out of school. It’s nice for recruiters to see any bootcamps, certificates, or online programs, though, because they show that you’ve furthered your education.
Although the education section of your résumé can give valuable information to the recruiter, you can’t really improve the section by doing anything less than going out and getting an additional degree or certificate, which we covered in chapter 3.
This section is where you can explicitly list all the relevant skills you have to contribute in a data science setting. Ideally, a recruiter will see this section and nod while saying “Yes, good,” because you’ll have skills listed that are relevant to the job. On data science résumés, there are two types of skills to list in this section:
Try not to list more than seven or eight skills to avoid overwhelming people, and don’t list skills that have no chance of being relevant for the job (such as an obscure academic programming language from your time in graduate school).
List only the skills that you’d be comfortable using on the job, not a language you haven’t touched in five years and don’t want to pick up again. If something is on your résumé, it’s fair game for a recruiter to ask about it. If the data science job postings that you’ve viewed request certain skills—skills that you have—make sure to list them! That information is exactly what recruiters look for.
We recommend that you don’t use star ratings, numbers, or other methods to try to indicate how strong you are in each skill. For one thing, ratings don’t mean anything: anyone could rate themselves 5/5. If you give yourself all perfect scores, hiring managers may think that you’re not honest or good at self-reflection; if you give yourself lower scores, they may doubt your abilities. Also, it’s not clear what you consider each level to be. Does 5/5 mean that you think you’re one of the best in the world, that you know how to do an advanced task, or that you’re better at certain skills than your co-workers? If a hiring manager does want you to self-assess your level of different skills, they’ll ask you in an interview.
Don’t list soft skills such as critical thinking and interpersonal skills; although they’re crucial to being a successful data scientist, putting them on your résumé is meaningless, because anyone can do that. If you do want to highlight your skills in these areas, talk about how you used them in specific instances within the experience section of your résumé. Also, you don’t need to list basic skills that anyone who applies would be expected to have, such as working with Microsoft Office Suite.
If you’ve done data science projects outside work, you can create a section for those projects. This section is great for candidates who have less work experience but have done projects on the side or in a school or bootcamp. You’re basically telling the recruiter this: “Although I may not have much relevant work experience, that doesn’t matter, because I’ve still done the full data science process.”
For each project, you’ll need a title, descriptions of what you did and how you did it, and the results. In fact, the data science projects should look as though they’re jobs in structure and content, so everything in section 6.1.2 on generating content applies to them too. Ideally, you’ll have a link to a blog post or at least to a GitHub repository that has an informative README file. Data science is a technical field in which it’s unusually easy just to show the work you did, and this section is a great place to do that. If you have enough relevant work experience, you can skip this section but still talk about projects in your interviews.
If you published papers that are related to data science in a master’s or PhD program, you should include them. If you published papers in other fields, even quantitative ones such as physics or computational biology, you can include them, but only briefly. Because they are not directly related to data science, the person reading them won’t get much out of the publications except that you worked hard enough to get published. You can list the relevant work you did during your research in the experience section, such as “created an algorithm to analyze millions of RNA sequences per minute.” But publication in a journal that the hiring manager has never heard of, even if it’s a prestigious one in your field, won’t go too far on its own.
You can add other sections, such as Honors and Awards if you’ve won Kaggle competitions or received a scholarship or fellowship, but they aren’t necessary. You don’t need to include references; speaking to your references will come later in the process, and you can share that information if you progress that far. Objective statements usually aren’t needed and are redundant, given the other information in your résumé. The phrase “data scientist experienced in Python looking for a position to develop A/B testing and modeling skills,” for example, isn’t going to make a recruiter more excited!
Generally, you put your contact information at the top, followed by the next-most-important section. If you’re in school or just graduated, you probably need to put your education at the top; if you don’t have relevant work or education, put your data science projects at the top; otherwise, put your work experience there. Within your work and education sections, list your experiences in reverse chronological order, from most recent to least.
We’ve seen lots of effective formats for data science résumés. In this field, you have a bit from freedom in your design; there’s nothing close to a standard format. Despite that freedom, you always want to focus on making your résumé easy to scan quickly. Because recruiters spend so little time looking at your résumé, you don’t want them to spend that time trying to figure out how to find your most recent job. Don’t make your design distract from your content; consider how others will view it. Some good practices include
If ideas such as whitespace and headers are overwhelming, stick to a résumé template you found online, or consult a design specialist.
Generally, you want to limit your résumé to a single page. This practice serves two purposes: given the brief skim your résumé will get, you want to make sure that the recruiter spends that time on the information you think is most valuable, and it shows that you can communicate concisely and understand what parts of your experience are most important to share. If a person submits a 17-page résumé (which we’ve seen), it strongly suggests that they have no idea what in their past makes them a good candidate and that they feel entitled to other people’s time to read it.
Finally, make sure that you’re consistent throughout your résumé. If you abbreviate month names in your education section, abbreviate them in your work experience section too. Although you can use different fonts and sizes for headings and body text, don’t switch up the format from bullet point to bullet point. Use past tense for previous positions and present tense for a current one. These things show that you pay attention to the small details and (again) help readers process your content quickly, as they won’t be distracted by font or style changes. A single inconsistency is unlikely to cost you an interview, but sometimes, details make all the difference.
It’s essential to proofread your résumé! A few typos or grammatical mistakes may lead your application to the (metaphorical) trash bin. Why so harsh? When recruiters are sifting through hundreds of résumés, two kinds stand out: those that are clearly exceptional (rare) and those that are easy to eliminate. The latter kind need some rules of thumb, and in addition to résumés that clearly don’t meet the requirements, résumés with typos are an easy reason to eliminate an applicant. Data science jobs require paying attention to detail and checking your work; if you can’t do that when putting your best foot forward in an application, what does that fact suggest about your work? In addition to using the spell-check feature in your word processor, have at least one other person read your application carefully.
We hope that coming up with the dates and titles of your work and education history is easy enough. But how do you come up with those punchy bullet points to describe your work experience (or data science projects)?
The common mistake people make on their résumés is to create just a list of their job duties, such as “Generated reports for executives using SQL and Tableau” or “Taught calculus to three sections of 30 students.” There are two problems with this approach: it states only what you were responsible for, not what you accomplished or how you did it, and it may not be framed in a way that’s relevant to data science. For the previous two examples, you could describe the same work as “Automated the generation of sales forecasts reports for executives using Tableau and SQL, saving four hours of work each week” or “Taught calculus to 90 students, earning an average of 9.5/10 in student evaluations, with 85 percent of students getting a 4 or 5 on the BC Calculus AP Exam.”
As much as you can, you want to explain your experience in terms of skills that are transferrable to data science. Even if you haven’t worked in data science or analytics, was there any data that you did work with? Hiring managers are willing to consider experience outside data science roles as still relevant, but you have to explain why they should. If any of your work can conceivably be related to taking data and understanding it, you should put enormous effort into creating a concise story about what you did. Did you analyze 100GB of star data for an astrophysics PhD? Did you manage 30 Excel files to plan staffing for a bakery? Lots of activities involve using data to understand a problem.
Have you used tools such as Google Analytics, Excel, or Survey Monkey? Even if those tools may not be the ones the job is asking for, working with data of any type is relevant. What communication skills did you use? Did you explain technical or niche concepts, maybe in PhD research talks or to other parts of the business? If coming up with transferrable skills is difficult, don’t worry; the rest of the advice on writing better bullet points will still help. But if you haven’t done so already, you should think about how your education or side projects can demonstrate data science skills, especially if your work experience can’t.
For the least-relevant positions that you held a few years ago, it’s okay to have just one bullet point. But you generally don’t want to leave a job off your résumé if it will leave a gap of more than a few months. If you’ve been in the workforce for a while and had a lot of jobs, it’s okay to list only the three or four most recent.
You might be finding that this process is a lot easier for the job you’re currently in than the one you had five years ago. One good practice is to keep a list of your accomplishments and the major projects you’ve worked on. When you’re in a job each day, making incremental progress, you can forget how impressive the whole is when you step back. People know that your résumé isn’t an exhaustive list, so they won’t think “It took her 15 months to build an automated system for tracking and scoring sales leads that saved their sales team more than 20 hours of manual work a week.” They’ll think “Wow, we need a system like that!”
In general, bullet points can fall into two categories. The first is big accomplishments, such as “Created a dashboard to monitor all running experiments and conduct power calculations.” The second category is average or totals, such as “Implemented and analyzed more than 60 experiments, resulting in more than $30 million in additional revenue.”
In either case, each bullet point should start with a verb and (ideally) be quantifiable. Rather than say, “I made presentations for clients,” write “Created more than 20 presentations for Fortune 500 executives.” It’s even better if you can quantify the impact you had. Writing “Ran 20 A/B tests on email campaigns, resulting in a 35% increase in click rate and 5% increase in attributed sales overall” is much more powerful than “Ran 20 A/B tests on email campaigns.”
Although the purpose of a résumé is to give hiring managers relevant facts about your work experience and education, the purpose of the cover letter is to help them understand who you are as a person. Your cover letter is where you can explain how you’ve researched the company and highlight why you’re a great fit. If your résumé doesn’t show a linear path, a cover letter can pull everything together and explain how the pieces fit to make you a great candidate for this job. Even just showing that you know what the company is, that you’ve read its About web page, or that you’ve used its product (if it’s available for individuals) goes a long way. Your cover letter is your best tool to help hiring managers understand things that don’t fit well in bullet lists.
Unlike a résumé, a cover letter may be optional. But if a company has a place to submit one, do so; some companies will eliminate candidates if they haven’t written one. It’s not uncommon for companies to give a specific thing for you to write about, such as your favorite supervised learning technique. This request usually is made to check whether people have read and followed the request of the job description instead of sending a generic cover letter everywhere. You definitely want to let the company know that you can follow instructions.
Knowing that a cover letter is to help the company better understand who you are, a common mistake we see in cover letters is focusing on what the company can do for you. Don’t say, “This would be a great step for my career.” A hiring manager’s job is not to help as many careers as possible; it’s to hire people who can help the company. Show them how you can do that. Even if this job would be your first data science job, what relevant experience do you have? What record of achieving results (even if they’re not related to data science) can you share so that it’s clear you work hard and accomplish goals? Don’t undercut yourself; try to think broadly about how you can make yourself appealing to the company.
Like your résumé, your cover letter should be short; three-quarters to one page is usually the rule. Focus on your strengths. If the job description lists four skills, and you excel in two, talk about those two! Don’t feel that you have to make excuses for skills that you lack.
Figure 6.2 shows an example cover letter.
Cover letters have a less-well-defined set of rules than résumés. That being said, here’s a good general structure you can follow:
The previous two sections lay out general rules for writing an effective cover letter and résumé. But the best way to differentiate yourself from other candidates is to tailor those documents to the position you’re applying for.
The first person who screens your data science résumé isn’t likely to be the manager for a position; it may not even be a human! At larger companies, applicant tracking systems automatically screen résumés for keywords, surfacing those that contain those words. Such a system may not recognize “linear modeling” as meeting the requirement for experience in “regression.” A human reader may not, either; a human-resources person may have been given nothing besides the job description and instructions to find promising candidates. You don’t want to risk a recruiter’s not understanding that your project using “k-nearest neighbors” means that you have experience in clustering analysis or that NLP is the acronym for natural language processing. You want someone to be able to look back and forth between your résumé and the job description easily, finding exact matches for the requirements in your experience. Although you don’t want to overload your résumé with tech jargon, you do want to use the keywords (such as R or Python) a few times to help the résumé make it through these screens.
We recommend that you have a “master” résumé and cover letter you can pull from rather than starting from scratch each time. This approach is especially helpful if you’re applying for different types of positions. If some jobs emphasize machine learning and others exploratory analyses, it’s much easier if you have bullet points and related key terms ready to go. Your master résumé and cover letter can be longer than one page, but make sure that the résumé and cover letter you submit are always less than a page.
Tailoring your application to the position doesn’t mean that you need to have one bullet point or skill for every single requirement. As we discussed in chapter 5, job descriptions are generally wish lists; try to figure out which ones list the core skills for the job. Sometimes, companies helpfully divide skills and experience into “requirements” and “nice-to-haves,” but even if they don’t, you may be able to tell which is which from the description of the job responsibilities. Although companies would love to get someone who gets a check-plus on everything, most won’t be holding out for it.
One exception is big tech firms and well-known, fast-growing startups. These companies get a lot of candidates and are looking for reasons to reject people. They’re very worried about false positives, meaning hiring someone who is bad or even just average. They don’t really care about false negatives—not hiring someone who is great—because they have lots of great people in the pipeline. For these companies, you usually do have to meet 90% of the requirements, if not 100%.
Company websites and job boards all have a place where you can apply, sometimes with a click of a button if you’ve saved your résumé on the job board. Unfortunately, because it’s so easy to apply this way, your résumé often ends up in a pile of hundreds or even thousands of similar cold applications. That’s why we recommend not applying this way until you’ve exhausted other options. Reading job postings is a great way to get a feel for what kind of jobs are available, but the best way to get your foot in the door is to have someone hold it open for you.
You want to try to use the hidden back door to most companies: referrals. A referral means a current employee recommending someone for a position, usually by submitting that person’s application and information through a special system. Many companies offer referral bonuses, paying employees a couple thousand dollars if they refer someone who gets and accepts a job offer. Companies like people who are referred because they come pre-vetted: someone who already works at the company and (presumably) is doing well thinks that this person would be a good fit. Even if someone doesn’t formally refer you, being able to write in your cover letter “I discussed this position with [Star Employee X]” and have that person tell the hiring manager to look out for your résumé are huge benefits.
How do you find people who can refer you? Start by looking at LinkedIn to see whether you know anyone who works at a company you’re interested in. Even if you haven’t spoken to that person in a while, it’s perfectly fine to reach out with a polite message. Next, look for people who previously worked at the same company or went to the same school as you. You’re more likely to get a response to a cold message if you mention something you have in common. Finally, look for people who are second-degree contacts to see who you have in common. If you’re on good terms with any of your mutual connections, reach out to that person to see if they’d be willing to introduce you.
If you’re reaching out to a data scientist, take some time to learn about what they do. Do they have a blog, Twitter account, or GitHub repo where they’ve shared their work? Mark Meloon, a data scientist at ServiceNow, wrote in his blog post “Climbing the relationship ladder to get a data science job” (http://mng.bz/O95o) that the most effective messages are ones that combine a compliment about the content he’s published with a request to ask some more questions. This way, you’ll also avoid asking about things they’ve already publicly talked about and can focus on getting advice that you couldn’t find elsewhere.
Remember that it’s not only people in data science who can help you. Although other data scientists are best positioned to tell you what it’s like to work at their company, people in any position can refer you. If someone you know works at a company you want to apply to, reach out to them! At the very least, they can still offer you insight into the company culture.
In his blog post “Do you have time for a quick chat?” (http://mng.bz/YeaK), Trey Causey, a senior data science manager at Indeed.com, outlines some suggestions for effectively reaching out to someone you don’t know to talk about your project, job search, or career choice. By following these guidelines, you’ll be much more likely to get a response, have a productive meeting, and build a good foundation for a continuing relationship:
Here’s how Trey pulls that all together into a sample message:
“Hi, Trey. I read your blog post on data science interviews and was hoping I could buy you a coffee at Storyville in Pike Place this week to ask you a few questions about your post.
I’m currently interviewing, and the part about whiteboard coding was really interesting to me. I’d love to hear your thoughts on how to improve whiteboard
I’m currently interviewing, and the part about whiteboard coding was really interesting to me. I’d love to hear your thoughts on how to improve whiteboard coding questions and answers, as well as share some of my own experiences with these types of questions.
Could you spare 30 minutes sometime—say, Tuesday or Wednesday of next week? Thanks for writing the post!”
Kristen Kehrer is a data science instructor at University of California—Berkeley Extension, faculty member at Emeritus Institute of Management, and founder of Data Moves Me, LLC. Data Moves Me helps data science teams communicate machine learning model results to stakeholders so that the business can make decisions confidently. She holds an MS in applied statistics and is a co-author of the upcoming book Mothers of Data Science (self-published).
Oh, a million! I come from a blue-collar family, where my dad was a firefighter and my mom stayed at home, so I was never taught how to write a great résumé for industry. But I did okay by asking others for help when I was getting out of grad school. I also have always been the type to keep track of any new project I work on or anything interesting that I could add to my résumé. I wasn’t one of those people who’d go for two years without updating my résumé. More recently, my old company paid for a career coach when they laid me off. I got to learn all about résumé best practices and how to effectively position myself to land a great job.
I absolutely advise people to update their résumé often. Especially if you’ve been working at the same place for a while, it is very difficult to try and think about all the relevant things that you could add to your résumé. For example, I co-authored a couple posters in the healthcare industry that won awards. That’s not relevant to every position I apply for, but if I am applying to a position in healthcare, I’d want to be able to reference that research. If I didn’t keep track of it, I would not be able to remember who my co-authors were or what the title of the poster was.
So many things! One is the four-page résumé that still has that they were a swimming coach. Another is not realizing that the applicant tracking systems don’t parse certain things well. If people have icons or charts on their résumé, that’s going to come through as a blob on a lot of the older automated systems and may end up with you being automatically rejected. I also don’t like when people put, say, three stars for Python, because you’re not giving people any context, and whichever skill you’re putting two stars for, you’re saying that you’re not good at that thing.
I’m not obsessive about it. But almost all medium to large companies now use an applicant tracking system, and you want to be able to rank high in terms of matching keywords. If I saw things on a particular job description that matched things that I’ve done, but were maybe worded slightly differently, I’d just edit a couple words to match the verbiage that they’re using on their job description.
I tell people to optimize their résumé for the job they want, not the job they have. You don’t need to make a list of all the things that you’ve ever done. Instead, think about what you’ve done that you can reposition for data science. For example, if you’re a math teacher, you’ve been explaining technical or mathematical material to a nontechnical audience. Or maybe you worked on a project where, even though it wasn’t in analytics, you had to work cross-functionally across multiple teams. Overall, you want to be able to show that you’re able to solve problems, self-manage, communicate well, and achieve results. Finally, you can use side projects to highlight your technical chops and the initiative that you’re taking.
You need to start applying to data science jobs. Too many people just keeping taking online courses because they think they need to know a million things to become a data scientist, but the fact is, you’re going to start a job, and you’re still going to have more to learn. Even ten years in, I still have more to learn. By applying, you’ll get feedback from the market. If nobody responds to your résumé, maybe it’s that you’re not positioning yourself well, or maybe it’s that you don’t quite have the skills. Gather some feedback from a few people and then choose an area to focus on, such as being able to automate processes in Python. Work on that, add it to your résumé, and apply to more. You need to apply, get responses, and iterate and move forward until you get a job.