Data Science Certificates in 2020 (Do You Really Need One?)


One of the most common questions about data science on Quora is “What data science certificate should I get?” For students who don’t have a formal educational background in data science, getting some sort of certification might seem like an important step on the road to a data science career. 

Let’s get one thing clear up front: you do not need any kind of data science certificate to get a job in data science. You should choose your learning platform based on the skills it teaches, not the certificate it issues, because recruiters just don’t care much about any data science certification.

But that doesn’t mean certificates can’t be valuable! Let’s dive deeper:

Choosing a Data Science Certificate Program

Finding a program that offers a certificate is the easy part. There are many data-science-specific programs that offer certificates of completion. A quick google search will turn up dozens.

What can be more difficult is comparing all of these programs. You should take into account the return on investment for entering into any of these programs, and whether the certificate is worth it at the end.

Some things to compare when considering looking for a data science certification are: 

  • What you’ll learn in the program

  • The cost of the program

  • Any prerequisites or qualifications you’ll need

  • The time commitment required

  • Reviews from learners who’ve done the program

 Let’s take a look at some real-world programs with data science certificates to make some of these comparisons!

Cloudera University Data Analyst Course/Exam

What you’ll learn: The course is primarily focused on conducting data analysis using Apache products: Hadoop, Hive, and Impala. Some SQL is covered but it does not appear to cover any programming in Python or R.

Cost: In its on-demand version, the course costs $2,235 for 180 days of access. The certification exam costs an additional $295,  so assuming you pass the exam on the first try, the total program cost is $2,630.

Prerequisites: Some prior knowledge of SQL and of the Linux command line is required.

Time commitment: Varies. Estimating roughly, each section might take around 5-9 hours and there are about 15 different sections. This is a self-paced course, and users have access for 180 days.

Reviews: Third-party reviews for this program are difficult to find.


What you’ll learn: Dataquest offers four different career paths that cover the skills required for data analyst, data scientist, and data engineering careers. Specific skills covered vary by path, but topics include Python or R programming, SQL and PostgreSQL, probability and statistics, machine learning, and workflow skills like Git, the command line (bash/shell), and more.

Cost: An annual Premium subscription costs $588 $294 (on sale). Monthly subscriptions are also available.

Prerequisites: None. There is no application process (anyone can sign up and start learning) and no prior knowledge of statistics or programming is required.

Time commitment: Varies. Dataquest is a self-serve interactive learning platform. Most learners find they’re able to meet their learning goals in roughly six months of studying fewer than ten hours each week.


IBM Data Science Professional Certificate

What you’ll learn: This Coursera-based program covers Python and SQL, including some machine learning skills with Python.

Cost: A Coursera subscription, which is required, costs $39/month. Based on Coursera’s 10-month completion estimate, the approximate total program cost is $390. A similar program is also available on EdX for $369.

Prerequisites: None. There is no application process (anyone can sign up and start learning) and no prior knowledge of statistics or programming is required.

Time commitment: Varies. Coursera suggests the average time to complete this certificate is ten months. 

Reviews: Quantitative third-party reviews are difficult to find, but:

  • 4.6/5 average on Coursera’s own site (112,020 ratings)

Harvard/EdX Professional Certificate in Data Science 

What you’ll learn: This EdX-based program covers R, including some machine learning skills, as well as some statistics and workflow skills. It does not appear to include SQL.

Cost: Currently on sale for $441.

Prerequisites: None. There is no application process (anyone can sign up and start learning) and no prior knowledge of statistics or programming is required.

Time commitment: One year and five months. Course progress doesn’t carry over from session to session, so it could require more time if you’re not able to complete a course within its course run.

Reviews: Quantitative third-party reviews are difficult to find, but:

These are just a few examples, of course! There are many certification programs out there, but you should research each of these five data points for any paid course you’re considering to ensure you’ve found something that’s going to work for you. 

Be Careful of Prerequisites and Qualifications!

While some programs like Dataquest, Coursera, and Udemy do not require any particular background or industry knowledge, many other DASCA’s Senior Data Scientist Certification and Cloudera’s CCA Data Analyst Certification have concrete prerequisites.

For instance, DASCA’s Senior Data Scientist Certification tracks require at least a Bachelor’s degree (some tracks require a Master’s degree) and a minimum 3-5 years of professional data-related experience! 

Some programs, particularly offline bootcamps, also require specific qualifications or have application processes, so you can’t just jump into learning. Consider the time costs and application fees of these sorts of programs when comparing them with other certification options.

Data Science Certificates and Your Job Search

Certificates certainly won’t hurt you in the job search as long as they’re presented correctly—see our Career Guide chapter on resumes for more details. But they’re unlikely to help much, either.

In the research for our Career Guide, we spoke with more than a dozen hiring managers and recruiters in data science about what they wanted to see in data scientist and data analyst job applicants. Not a single one of them mentioned certificates at all.

For a potential employer, the problem with data science certificates is that there’s no universal standard and no accepted certification authority. Different websites, schools, and online learning platforms all issue their own certificates, and assessment standards are generally low.

That means that at a glance, an employer seeing “Data Science Certificate” on a resume isn’t going to know what kind of training the site or school issuing that certificate offers. They also aren’t going to be able to tell whether the site or school in question has a rigid grading policy (almost none do), or whether it has an effective way of confirming online students’ identities (almost none do).

What About University Certificates?

Even for many of the expensive certification programs offered online by brand-name schools, including Ivy League schools, certificates are not very meaningful to potential employers. Many of these online programs are not actually administered by the schools themselves; they’re run by for-profit third party firms called Online Program Managers.

Moreover, the ways students are assessed in an online certificate program can vary dramatically from the way real-world students are addressed in normal university classes. Employers are aware that a Harvard-affiliated certificate from EdX, for example, and a Harvard University degree are very different things.

Most data science hiring managers don’t have the time to research the academic rigor of every data science certification they see on a resume. The average resume might only get 30 seconds of a recruiter’s attention, so rather than focusing on certificates that won’t tell them much about a candidate’s ability, they’re going to focus on the areas of a resume that will give them the info they need: skills and projects.

Entry-level applicants can be assessed more effectively by the data science projects they include on their resume and GitHub. Higher-level applicants will be assessed mostly on previous industry experience. It just doesn’t make sense for hiring managers to spend time trying to chase down whether a particular data science certificate means anything when the information they actually need—whether the student actually has the skills they need to do the job—is available elsewhere on the resume.

So What’s the Point of Certificates?

Data science certificates aren’t useless, of course! At Dataquest, we issue certificates for course completion, because we see it as a good way for some students to highlight that they’re actively engaged in learning new skills. Recruiters do like to see that applicants are constantly trying to improve themselves, so listing certificates can help your job application in that way.

Make no mistake, though: the effect is likely to be minor, if there’s any effect at all. What’s most important to recruiters is whether you can actually do the job, and that’s information they’re going to find primarily by looking at your portfolio of projects. They’re going to look at your projects and skills first, and either eliminate or pass your application to the next round based mostly on what they see there. Continued learning efforts are something you can and should bring up in an interview, but they’re very unlikely to get you an interview. 

Long story short: data science certificates are unlikely to help much in your job applications, particularly at the early pass/fail stage of resume assessment, so you shouldn’t focus on questions like “which data science certificate is best?” when you’re trying to figure out where to learn data science.

Find the platform that best helps you learn the crucial data science skills, because that’s what’s actually going to help you land jobs in the field.

The post Data Science Certificates in 2020 (Do You Really Need One?) appeared first on Dataquest.