What is resume parsing and how an applicant tracking system (ATS) reads a resume
“Anonymous employee at indeterminate company with an unspecified degree in something totally unidentifiable seeks an unrecognized job in the complete unknown.”
This is what your resume looks like to even the most swanky Applicant Tracking System (ATS). Unless you are applying to be a spy, this air of mystery is unlikely to work in your favour. In this post, I’ll explain how to write a solid resume that will pass muster with even the most stringent of parsers.
In my final year at university, I recall attending a seminar on how to write a good CV. This was back in the day when you thought you were smart to buy thick, expensive paper to print it on. The papyrus I chose would have shamed a Pharaoh and jammed most printers. I remember spending hours trying out different fonts and formats which, being an engineer, was not really my forte. Yet all the advice I ever got from books and seminars could be boiled down to one short phrase: make your CV stand out.
In those days it meant that your resume would have to make itself visible amid stacks of hundreds of other printed CVs weighing down some poor unfortunate desk and the person who sat behind it. Thankfully, hiring has changed. CVs are now stored digitally and increasingly fed through automated resume parsing recruitment software like Workable.
What is resume parsing?
A ‘parse resume’ definition we can use is ‘the process by which technology extracts data from resumes.’ This means that the job of the parser is to extract the key components of your CV, such as your name and email, the degrees you hold, the skills you have and your work experience. Which is pretty much what we are building with Workable. We’re very good at this but honest enough to admit that it’s hard. The facility for language of even a modern-day resume parser hasn’t yet reached human levels. In other words, you’re no longer penning a resume for someone who might prize quirkiness, Pharaonic paper or originality; you’re writing it for a parser, which wants you to follow standards.
It’s the text that counts
To parse resumes, you need to analyze text. Your CV, with its soft pastel font colours, fancy fonts and WordArt headings could be beautiful enough to have been illuminated by a monk but the chances are it will just look like noise to a parser. Parsing resumes involves text extraction which means exactly what it says on the tin: extract the text and ignore the rest. So what could possibly go wrong there you might think? Take a look at how our budding applicant, whom I shall refer to as John Doe, can have his name lost in translation:
J O H N D O E
Did you notice that I left an extra blank line at the top there? It’s not a typo folks, that’s how John Doe’s name appears after text extraction when John Doe creates a custom banner in Photoshop with his name and contact details (also among the dearly digitally departed) and pastes the image into his CV. Large headings, unusual character spacing and font choice can result in spaces being created or lost, or characters being repeated as the next few examples show. Tables and columns will put words, and sadly sometimes letters, on different lines. So, you need to follow some basic rules to make sure your text can be extracted properly:
- Submit your CV in a text format preferably .doc or .docx. There are plenty of open source word processors that understand this format these days
- If you use the PDF format, make sure you export it from your word processor – don’t scan your CV into an image.
- Try to avoid using headers and footers as they often get interspersed with the main body of text
- Use one standard font throughout the CV
- Don’t use tables and columns as the ordering of sentences may not be what you expect
- Don’t use WordArt
- Don’t fiddle with character spacing
- Write your document on your own computer so that your metadata is correctly set. Text can be found in there too.
- Put your name in the filename of your CV
This list may sound overly strict and at the end of the day, you do want to present a document that is well formatted, tidy and looks professional. I would argue that you don’t need to use any of these features in order to achieve your goal. I’ve painted a somewhat bleak picture here and the truth is that in some cases errors occur in places you don’t really care about, and in others the errors themselves can be overcome through intelligent resume parser design. The key point is, why take a risk on something so important? And remember, the key thing is that it’s what you have actually done that matters the most.
Don’t take my word for it, take a look at what happens behind the scenes and judge for yourself: With the text extracted, the parser’s next task is to look for words and phrases that it would expect to find in a resume. Artificial intelligence has not yet advanced to the point where a computer can interpret text at anywhere near the level that a human can, they do have an amazing way of remembering a vast number of things: names, job titles, companies, countries and cities are just some examples that a parser retains a deep knowledge of. Unless of course your job title is a freshly-minted neologism that sounds more like a Chuck Norris movie title and less like a job title.
Following standards is not just about job titles and text layout. Section titles are important too, as are the following standards I would recommend following:
- Stick to a chronological resume format, not a functional one
- Use typical names for section titles like “Education”, “Work Experience”, “Personal Details” etc.
- Use a date format appropriate to the country you are applying in and make sure you include dates in full (day, month and year) to make them easily identifiable
- Only use common and well known abbreviations such as CTO, MBA etc…
- Use typical names for job titles and avoid fancy embellishments
- Use a spellchecker. Typos and misspellings make you look bad to resume parsers and humans alike.
But I’m creative!
I can already hear the cries of graphic designers, whose resume often doubles up as a canvas for their creativity, a platform for making a statement of self-expression and uniqueness. If you have a legitimate reason to need a fancier looking CV, my advice to you is to maintain two versions, one of which is ATS-friendly. When you next apply for a job, find out which version will be more appropriate.
Next time we’ll be taking this advice and putting it to practice as we’ll present some CV templates for your candidates that will pass through a resume parser with flying colours (but I assure you they’ll be in monochrome).
Related: How to maximize user adoption of your ATS
Frequently asked questions
What is parsing in ATS?
Parsing Technology allows recruiters to electronically gather, store, and analyze large quantities of resumes. Once acquired, the data can be easily searched through for successful matches that are perfect fits on your hiring needs.
What is parsing in recruitment?
The process of turning your resume into a computer file that job seekers can use to apply for jobs. This is done with various software tools, such as the parsers in this article's title - they're designed specifically for looking through resumes and extracting key data about you so employers know what kind of person would be a good fit with their company culture.
What is the purpose of ATS?
An applicant tracking system (ATS) is a great way to keep track of all your potential hires, from initial application through onboarding.