115

Lately I've come across a lot of resume, portfolio or business websites showing the level of proficiency in a certain array of skills by using percentages.

For example, I saw something like this on the website of a small company involved in web design:

  • HTML : 80%
  • CSS : 75%
  • Javascript : 50%
  • PHP : 30%

This confuses me. Does this mean you master something when you reach 100%? Is it possible to have 0% in a skill? How much XP do you need to gain 1 more percent?

21
  • 111
    Upvote for the "How much XP do you need to gain" comment.
    – JasonJ
    Jul 13, 2016 at 14:09
  • 92
    I once saw Reliability: 80%.
    – camden_kid
    Jul 13, 2016 at 15:55
  • 35
    This seems quite likely to fall victim to the Dunning Kruger Effect. If you don't have 100% mastery, how can you assess how close you are to 100%? (also, no one has 100% mastery)
    – Chris G
    Jul 13, 2016 at 17:02
  • 22
    @ChrisG What about the guys with 5 years of experience for a 3 year old language that everyone wants? I bet they have like 160% mastery.
    – corsiKa
    Jul 13, 2016 at 17:40
  • 29
    @ChrisG You've obviously never heard of Jon Skeet. Jul 13, 2016 at 19:02

8 Answers 8

40

The percentage scale makes sense if the company administers a pre-interview technical test and these are pass marks.

Otherwise the percentages don't make sense. After all a percentage only makes sense in reference to something else.

What would make more sense is:

  • Prescribing number of years' experience
  • Asking for specific qualifications

If you are applying to a job advert with this information, I would try to address your competencies on each of these areas citing your experience, qualifications and (most importantly) what benefits you can provide to an employer.

11
  • 3
    I see. I was wondering whether or not using that kind of list for my web resume would be a good thing. I think I'll pass.
    – Dryr
    Jul 13, 2016 at 14:14
  • 25
    A very old problem with years experience is it doesn't detail depth of experience. 10 years of experience could mean doing the same thing for 10 years or it could mean 10 years of progressively more advanced experience. Jul 13, 2016 at 17:25
  • 7
    Also typically jobs involve multiple skills. If you are a full stack web developer using C#, JavaScript, SQL, HTML, CSS for 2 years do you claim 2 years for all of them? Likely your skill level at any individual one will be less than someone specialising in one for 2 years. Jul 13, 2016 at 18:52
  • 7
    @RichardU or it could mean you learned the skill 10 years ago and have used it only sporadically since. Jul 13, 2016 at 21:24
  • 2
    @RichardU: Yep. I've worked with someone who purportedly has 40 years experience, but certainly no actual aptitude. Jul 14, 2016 at 14:48
16

Why describe the level of proficiency in a skill by using percentages?

This allows for lazy decision making by people doing initial screening. Remember that numbers are magical to some people so they believe that PHP=40% is better than PHP=30% even if those numbers are arbitrarily self assigned.

5
  • And if you are not a lazy decision maker, you may want to use the presence of percentages as an indication of whether someone is interested in clearly presenting reality or playing tricks. =)
    – jpmc26
    Jul 14, 2016 at 4:59
  • @jpmc26 Advising users to do this is becoming common, so presence of percentages for skills wouldn't be a solid indication of people playing tricks.
    – Myles
    Jul 14, 2016 at 14:12
  • 2
    What nimrod is advising this? =/ (As in, what group of at best poorly informed people.)
    – jpmc26
    Jul 14, 2016 at 15:36
  • 4
    @jpmc26: maybe the same kind of person who expects an answer to the question, "how much does it hurt, on a scale of 1 to 10?". They don't take kindly to being asked whether they mean an arithmetic or logarithmic scale from 1 to 10, let me tell you. Jul 15, 2016 at 8:15
  • Sadly, this answer is probably the closest to the truth.
    – Mast
    Jul 15, 2016 at 10:52
6

I agree, percentages like this are useless. Without some set standard of what 100% is me saying 78.8% is just a arbitrary number.

As to what you can do about it? If you need to rank your own skills, ask first what they think 100% is, and when someone gives you a percentage, ask them what 100% means to them.

4
  • Good point. If an employer needs this kind of grade, I'll ask what the base values mean.
    – Dryr
    Jul 13, 2016 at 14:21
  • 2
    Hah! my skill is infinity +1 Well, actually that makes as much sense as any other rating. I agree it's laziness but if I were forced to use such a tool and I was a hiring manager, I'd exclude anyone rating themselves over 90% because either they're lying, or are more skilled than I need Jul 13, 2016 at 17:28
  • @RichardU no need to lie here, just not being competent enough to grasp the extent of their incompetence.
    – Chieron
    Jul 14, 2016 at 8:14
  • 2
    @Chieron you'd not believe the levels of incompetence I've dealt with over the years... Hearing "What is the difference between a sub and a function" from someone who is allegedly a VBA developer is enough to question your own very existence. Jul 14, 2016 at 12:11
5

These are just numeric representations of vague statements. Using the sample values:

  • HTML : 80%
  • CSS : 75%
  • Javascript : 50%
  • PHP : 30%

Could mean:

  • I know what HTML is. I work with common HTML tags, like A and BR and SPAN and DIV... I don't have it all memorized, though, so I may not be able to tell you that the BLOCKQUOTE tag has a CITE attribute (particularly without looking it up). I am more comfortable with CSS. Maybe I've mastered (X)HTML 4.01 but just haven't caught up with some of the HTML 5 changes.
  • I know how to create a stylesheet and know many of the common tags. However, I would say my CSS prowess isn't quite as good as my HTML skill.
  • I am familiar enough with JavaScript that I am not confused by it. I can do some things.
  • I know PHP well enough to recognize that the code is PHP. I may be able to make some simple changes to code without breaking stuff. I would probably be better off with a help guide.

All of this is basic conjecture. As you can tell by glancing at the other answers, there is no real formal process for precisely calculating the numbers. So what this means is that we must take these numbers as rough estimates, and possibly guesses. * I have been known to spend roughly 80% of a standard-length work day in HTML, but only 30% of my career was doing PHP-related stuff.

We don't know, just based off of the numbers, what 100% means. It may mean any one of these things:

  • 100% mastery of the every common part of the language and many uncommon parts
  • 100% ability to understand the language and understand every detail with quick scans of relevant documentation
  • 100% ability to be able to work on this professionally, alone
  • a high likelihood to get 100% on an exam (college level exam for an entry-level class, or entry-level industry certification)
  • length of a work day
  • length of a career
  • any other thing that the person might be thinking as "100%"

I can think of multiple reasons why a person uses numbers instead of clear sentences that have more meaning. One is that percentages may be much faster to read, which may be a benefit.

On the other hand, rather than trying to respect the time of a hiring staff, the person might be intentionally being a bit vague, so that specific and verifiable claims don't go on record in a way that could be the basis of taking formal/legal action against them (such as terminating employment on the basis of provably false statements on the resume).

Another possibility is that percentages may show some relative strengths, just trying to show HTML is better than PHP. (In fact, with many people these days, they might even be forgetting that 80% means four fifths of something, and might just be giving themselves scores using numbers that they see more frequently, like progress in a video game.)

Ultimately, this is failing to communicate with some people. One of the key principles of successful communication is that the message must be successfully understood, not just transmitted/delivered. Proof that this method is failing can be seen by looking at the chorus of other answers to this question. So far, they aren't offering anything more specific either. However, the author may not be realizing how much struggle people have with interpreting whatever they are intending to communicate.

So, don't do what they did. Don't provide meaningless numbers (without any other details that provide sufficient context to understand them).

8
  • 27
    HTML : 80% could also mean that I know what H, T, and M mean, but I'm not so confident of L.
    – Peter K.
    Jul 13, 2016 at 19:33
  • 5
    @PeterK. But that would be HTML: 75%
    – simbabque
    Jul 14, 2016 at 14:13
  • 9
    @simbabque ah, he's more clever than you think. HTML 75% would mean ZERO knowledge of "L" where his example gives 5% which is 20% of the remaining % so that is enough to be unsure :D Jul 14, 2016 at 15:02
  • @RichardU maybe he should be in financial politics instead of web development instead.
    – simbabque
    Jul 14, 2016 at 15:10
  • 5
    @RichardU : What makes this particularly amusing is that web pages content can commonly be found in files that have names ending with ".html" as well as files that end with ".htm" so there is actually a widespread difference in whether people prefer that L.
    – TOOGAM
    Jul 14, 2016 at 15:10
4

As you already pointed out, obviously this kind of presentation is somewhat relative and therefor meaningless to directly compare job applicants.

In my opinion, the only usefull information which you can transport by categorizing your skills into percentages, is the relative weighting which you give to the particular skill in comparison to the other skills listed.

90% PHP
30% CSS

Would therefor have the meaning, that you have a strong focus on PHP development. It wouldn't necessarily mean you are an PHP expert.

As you said that you saw this a lot on portfolio or business websites, I must point out that a lot of sites display this info in excactly that way, because many stock themes feature such a presentation (here is a typical example). A lot of small companies use stock themes to build their website nowadays, because it's cheap and fast.

Such websites are build design first and content last - which means that in the majority of the cases, there hasn't been put much thought into the relevance of the information for the visitor itself (it only has to look good ...).

I would only take it as a hint, showing the technological emphasis of that company.

I have personally never seen it in a resume and also wouldn't advise you to use this in a resume.

1
  • 3
    Exactly, it is mainly helpful relative to each other as opposed to the mastery as a whole. I've "learned" probably a dozen programming languages to varying levels, yet stuck with web languages for the last 4 years. If I was applying to a new job I wouldn't want to put Javascript, which I consider myself an expert, and Java, which I remember only the basics, as weighted the same on my resume. It would be misleading. Assigning them percentages gives a self assessed weight to them.
    – DasBeasto
    Jul 13, 2016 at 18:14
3

I recently had this in an interview where I was asked to rate my knowledge and/or proficiency in a variety of things and after which I then gave my explanation for why I chose that particular rating. Afterwards when I was given some feedback about it, they mentioned that in one of the questions I answered with a low rating and gave my explanation why I chose it and another candidate gave themselves a near perfect rating using the exact same explanation I gave for my poor rating!

In essence it seemed like they were using it as a method not only to gauge how experienced the candidate was (through further prompting) but to gauge how accurate the candidate's perception was of their own abilities. I think when it's used in this context it can be quite beneficial, but as others here have mentioned if a rating of any kind is used without some time of criteria to compare it to - it's useless due to its subjectivity - As my example above clearly illustrates - another candidate and I both described ourselves as having the same skills in a certain area - but one came off with a very high view of their ability, and the other didn't.

4
  • But this would just lead to more questions. Is his high view of his ability due to delusions of adequacy? Or confidence in his ability to learn? If so, is that confidence correctly placed (i.e. can he learn more about it easily, building on the framework of knowledge he's developed?) Conversely, is your low view of your ability due to self-doubt, or realistic assessment? Is it based on knowing there is more you can learn, or is it based on believing there is much more to that subject that you will never be able to master? This whole approach seems fairly negative and extremely subjective.
    – Wildcard
    Jul 14, 2016 at 23:24
  • @Wildcard, I think you raise good points. I can see how they can potentially be useful if used along with the right follow up, but as you rightly pointed out even if you know that their perception is inaccurate; you don't necessarily know what's the cause behind that (without proper prompting) and that in of itself is important too.
    – fib112358
    Jul 15, 2016 at 7:01
  • 1
    @Wildcard: it could even be that they both have realistic views of themselves, but a different perception of what a top score means. If the other candidate took 100% to mean "meets the requirements of this job" and fib took it to mean "an Olympic champion" then obviously they come up with different rating for themselves. I would suspect this question of being designed to filter out candidates (like myself) who genuinely struggle with a mental block when asked to assign utterly meaningless numbers to unquantified properties ;-p Jul 15, 2016 at 8:21
  • For what it's worth, the question was something along the lines of "How would you rate your knowledge of Linux?" and I answered with "I've installed a few distros, I am familiar with how to install programs from the GUI but I haven't really touched the terminal and I don't know how to shell script so I'd give myself a 3/10" and the other guy replied with the same answer (in a positive spin of course) and said 9/10 (this isn't what it actually was as I'm actually pretty familiar with Linux but it was the exact same spirit).
    – fib112358
    Jul 15, 2016 at 8:35
2

I am going to propose a possibility drastically different from every other previous answer.

If you say you see a lot of these crop up, there is a chance that they are a result of some sort of automated (probably online) test/quiz.

The ratings definitely match up to what they would be if it was an online quiz, many of which like to use percentage results.

Now, if that's the case, you can be the judge of just how accurate an online quiz is at assessing proficiency. I'd assume not very good, without proof to the contrary.

1
  • 1
    " I'd assume not very good, without proof to the contrary." -- but at least if you can figure out the (usual) curriculum for the kind of tests in question, then you have some idea what the number means. It may not reflect real-world proficiency with the technology, but training to a test is a proficiency of its own, so it's measuring that :-) If you're right about tests, then these numbers might mean for example, "has seen all of these things before, prepared for the HTML and CSS tests, half-heartedly crammed Javascript the night before, got a few PHP questions right on general knowledge". Jul 15, 2016 at 8:28
1

The numbers are intended to give the candidate's self-assessment of various areas of expertise. They're to give a starting point for the conversation.

I've been asked by technical recruiters to rate my various skills in a range from 1 to 10. How is your level of X expertise? I haven't coded in X in years, am not interested in X and am not advertising my skills in X. Answer: 2 of 10.

How about language Y? I've been coding in language Y for 5+ years. I've been involved in the SO community for that language for 4 years. I am fairly confident in my ability, but I'm not John Skeet. 8 of 10.

As an interviewer, you use those numbers no differently that any other information available on the resume. You need a DBA and they rate themselves as a 1 of 10. Are they willing to learn and you willing to train? You may have a match. They rate themselves a 9 of 10 and you're looking for a 1? Maybe look at someone else.

Don't think those numbers are reliable? Great! You're a smart interviewer. Your job is to test the candidate and see if they're right for the role. Those percentages are no more or less misleading, unhelpful or dishonest than anything else listed on any resume. Caveat Emptor.

2
  • There are a couple points on resumes that are worth reading. Only a couple—and they're the key ones that are worth writing, as well. I read a great article about this recently.
    – Wildcard
    Jul 14, 2016 at 23:30
  • So many people fall victim to the Dunning Kruger Effect and misidentify their own skills, or they are aware of the DK Effect and try to account for it, and may therefore give answers that the interviewer didn't expect. I agree with a lot of the posters here that this sort of measurement system really only works well when specific skills tests are being given and score targets are set for various jobs (e.g. must score 30% on Java test to be hired as Jr. Java Trainee Sub-Intern Bit Bucket Janitor, must score 80% to be hired as Sr. Java Developer), but I haven't found many jobs like that. Aug 28, 2016 at 1:06

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .