Friday, May 28, 2021

How HR Computer Programs Read Your CV


If you send an application to a major company today, it is very likely that a computer will receive it there first.  More than 70 percent of companies already use so-called applicant management systems. These scan applications for predefined terms and match them with the job ad.  Do you have the degree required by the ad?  If not, the application may end up in the trash before a human employee has read it.

But not only then.  Often, minor mistakes are enough - and your application is sorted out.

HR departments don't have to compare several hundred applications by hand, and applicants get clarity faster. But the programs have their pitfalls and sometimes fail, for example, when it comes to formatting a resume.

How do you write a good resume?

To test the documents so-called CV parsers are used.  The program scans resumes for hiring criteria and specified search terms.  Then it generates a report on what it (didn't) find.  This allows people to spot problems before they send in their documents.

What exactly does a CV parser do?

Normally, the respective department reports the desired profile for new employees to the HR department.  Based on this, the HR department creates the job advertisement and enters all the important criteria into its applicant management system.   For example, final grade, professional experience, or field of study.  The CV parser scans all applications received for these criteria and pre-sorts them in a database.  Information that the program does not recognize is not transferred.

The database is later used by the HR and specialist departments for keyword searches ("Show all applicants with relevant work experience / the right degree / the right grade") - sometimes employees don't even look at the CVs themselves.  Thus, a lack of machine readability can become a pitfall for the application.

An Example:

Here is a fictitious applicant: Christine Sanders. After graduating from high school in Augsburg, she completed a degree in business administration with a focus on textile retailing. At the same time, she gained professional experience at several large textile chains and is now doing a traineeship for junior managers.

Carla Aubin checked Christine's fictitious resume for errors using her CV parser.  The result: the program had problems understanding the information on seven points - and what the computer didn't understand, it either didn't pass on or sorted incorrectly.

Studies and final grade

Christine took a specialized business administration course, but the term "business administration" does not appear.  So if the company had specified a business degree as an exclusion criterion, Christine would have been automatically sorted out.  Applicants need to use the term mentioned in the job advertisement as such.

Also relevant: Christine completed her bachelor's degree with a grade of 2.0, and her high school diploma with a grade of 1.7 - any human employee in a human resources department would probably recognize that.  However, the CV parser had problems with this because the grade was not preceded by the word "degree".  Because the grade is listed on the same line as the degree program.  If she had instead written "Final grade: 2.0" on a separate line, the program would have been able to read and match the information.

Foreign languages or driver's licenses?

Very good English, even better French, and basics in Swedish: Some parsers, however, recognize the combination of letters and numbers standardized for language levels as a driver's license - or not at all.  In the test, the program made an upper level in English and did not even show French and Swedish in the evaluation.  Probably because they are listed on the same line as English.

Better use the terms 'fluent' or 'business fluent'. In addition, he says, you should make sure to list all the languages required in the job posting - which means also noting that you have a command of German, even if you can assume that if you have a German high school diploma.  CV parsers are usually not proficient in transfer skills: in the evaluation of Christine's CV, it was stated that she was not proficient in German...

Artistic elements

More and more CVs are sent in with graphical elements, especially in creative industries: Bar charts, diagrams, or star ratings reminiscent of Google are supposed to make clear at a glance what knowledge one has.  However, CV parsers usually don't understand these images at all.  A simple listing will get you further in this case.

Date formats

Especially when applying to multinational companies, you can get into trouble because of the dates. The 4th of July this year would be written "7/4/21" in the USA, in Germany "4.7.2021". Ricci also couldn't keep track of exactly when he worked at all of his jobs. As a result, he alternates between year dates, month dates, and day dates in his resume.

In the test, this led to the fact that, in the parser's eyes, Ricci had been in a student internship at a cabinetmaker's shop continuously since 2013. "Ideally, therefore, you should always use only the format MM.YYYY when specifying stations in your resume," advises Dehn.


Christine comes from the textile trade, DOB is a common term there for "women's outerwear," AWS stands for "accessories, lingerie, socks."  Specialists would understand this, but the program doesn't - and sorts Ricci's professional experience as PC knowledge instead.  So it is better to write out all terms.

Change or promotion within the company

The CV parser usually recognizes only one of them.  "Information from applicants who have held two or three different positions within a company and deal with that in one CV station is usually never well understood by the program.  Better to list the different tasks as individual items.

Everything on one page

As a service to the recruiter, applicants like to try to keep the resume concise and compress it into two columns on one page.  This is one of the biggest problems for the computer.  Parsers often scan from the far left to the far right; unrelated information is thus quickly mixed up.  If you abandon the romantic notion that a person reads the application first, you can submit a multi-page resume without hesitation.

Convincing with the letter of motivation

It is true that the human eye understands and corrects many such errors during matching. But parsers are used precisely to save time.  Often, therefore, the original is no longer even looked at.

And what about the cover letter - is it still important today if the application is evaluated by a computer anyway?  Yes!  But especially in the second step. Because once you have overcome the hurdles of the algorithms, a human still reads the application.  And they still want to be impressed with a good cover letter.




No comments:

Post a Comment

Note: Only a member of this blog may post a comment.