An experiment in A/B Testing my Résumé | Hacker News Reader

An experiment in A/B Testing my Résumé

paulbutler.org

109 pointschanux27 comments

Having screened many, many resumes from co-ops and interns, I buy the shorter resume arguments. A lot of people put crap that some job counsellor told you will make you look more rounded, but the sad truth is that I care about your ability to not mess up the code base so much more than the fact that you volunteered on 'campus day'.

It also makes you stand out in a very basic way. In the stack of 30 resumes, you'll be the one person who could articulate in 1 page where everyone else took 2.

How many two page resumes do you get? I'm in college, and I've never seen one of my peers take a two page resume to the career fairs.

I've read TONS of two page resumes. I've noticed that headhunters seem to encourage them. However, I'm not going to spend more than a few seconds looking at any given resume, so with multiple pages the chance that I'm going to see what you want me to see is far lower than if it were condensed down.

Really hard to fit years of relevant experience on a single page; you'll see a lot of two page resumes from people with more than a few years of experience. Of course college students don't have two page resumes, they don't yet have enough material that actually needs to be on a resume: they typically have a skills section, an education section, one or two short-term relevant jobs, and maybe enough fluff to fill the rest of the page.

In Germany, Résumés aren't necessarily limited to a a page. Two pages is considered fine too. Three pages is where it starts to be a problem :) Somehow related: The european union standardized a format for a CV:

https://europass.cedefop.europa.eu/europass/home/vernav/Euro...

(Which is a horrible idea imho, but still...)

I haven't screened too many resumes but I prefer to see everything. I'm in IT, not HR so it's not like I have 5,000 other resumes to go through per day. The person working for me will be someone I work with all the time and I don't mind spending 35 minutes instead of 15 to go through the resumes.

Show me everything you got. Show me the weird but cool techy projects you've been involved in. Show me the platforms/technologies you've worked on and show me the clever stuff you did. Tell me about the role you played in the team and your personal contribution to the project. As far as I'm concerned saying "added mDNS support to project" beats "lowered TCO by $1.5m."

I don't care about action words or acronyms. I care about what you actually did. "Actuated exemplary leadership skills" or "Decentralized payroll processing system" doesn't tell me anything. Tell me you wrote a C-module to interface with backend-DB or scripted rsync to remotely backup 75 Windows desktops. Basically, tell me you know what you're talking about by giving me an insight into your career not a mish-mash of big words, acronyms, and jargons.

Also, customize your cover letter. I actually do read it because I want to work with you for a long time.

Customizing resumes or websites for the audience is the logical next step after A/B testing.

I've always customized my resume for the job I'm applying to - isn't that common?

It's common sense, but unfortunately it's not common :(

It also discourages people from embellishing their resume with stuff that they don't really know.

I think a lot of especially younger people don't realize that if you list a programming language on your resume it's sort of expected that you'll be able to write something in that language (maybe with minor syntax errors). I'm appalled at the number of people who list C/C++ and yet can't write a simple string manipulation function.

Actually, this phenomenon is kind of useful because it the quickest way to filter out more than half of the candidates applying.

From my experience, a lot of employers are Nash level-two thinkers here: they know that everyone embellishes their resume, so they assume that you've put union of the sets "things I know how to do" and "things you require" on it, rather than their intersection. Thus, they require "10 years experience in Clojure" etc. because they think that anyone who doesn't say they have it really doesn't know anything about it, even though that usually disqualifies the Nash level-three candidates who honestly evaluate themselves.

In France, résumé is only one page because the résumé will give you an interview. And the interview will give you the job.

If I could, I'd hire the guy based that blog post alone.

Sorry, Facebook already did.

The stats are a little daring, I wouldn't consider the social media link effects interesting ("significant") at all. Short versus long is a good thing to back up with some data, though.

Could the fact that his "conversion rate" went down when people were given outbound links be explained by the fact that they would rather visit those links than keep reading? I know that if I saw a resume with both a link to a GitHub profile, and to a blog, I'd ignore the blog, stop reading the resume (though perhaps downloading it first in case I needed to contact them) and start reading their code.

also, if the reader is a recruiter, accessing the linkedin profile may be more useful.

I'm surprised at the comment about wanting to drive clicks to his blog rather than other sites (namely Github).

When I'm interviewing someone, I am far more interested in a github link thank a blog. Everyone has a blog and most aren't very interesting. But their github account.... if I am hiring a developer, I'll be very interested in what code they've written, projects they're working with, etc.

Looking quickly at both, his blog is more interesting than his GitHub page, I think. Also he can track traffic to the blog and every page of the blog has a link to his GitHub page anyways.

Precisely. I use GitHub as a tool and my behavior would change if I started using it as a portfolio.

I've heard from dozens of recruiters through my blog, but I can't remember ever being contacted for work from my GitHub.

I also have pretty specific interests (data and data science). I'm pretty picky about the type of work I'll do, so I'd rather be hired for the stuff I talk about on my blog than my GitHub.

Does he mention how many hits he actually got in total? I can't seem to find it in the post

I didn't, because I vaguely remembered reading that discussing CPC could get your account terminated. Now that I look at it, it seems like that applies to AdSense, not AdWords.

I did this a while ago so I don't remember the exact number, but it was on the order of 300 clicks.

I'm glad to see the longer/shorter resume results are consistent with the first piece of advise given by Manager Tools on resumes: "one page". http://www.manager-tools.com/2005/10/your-resume-stinks

I've followed this advice and found that most recruiters and employers love a one-page resume. A few hate it, but that's fine by me: I'd take love/hate over indifference any day.

He seems to miss calculating the total visits to his blog based on links included. Seeings as it looks like 5x the number of downloads with a LinkedIn link, but only 1/4 the blog view rate, he actually gets a higher absolute blog view rate (5/4 the number of people) as without a LinkedIn link. The effect is even more pronounced with GitHub (2x absolute number of views), although I agree he should leave Twitter off (1/3 the number of views).

The denominator for the "blog view rate" is the number of times the landing page was shown, not the number of downloads (the landing page showed the entire resume and had the same outgoing links.)

Interesting idea - Thanks for sharing your findings. I'm not sure how useful the data is though. You drive a lot of traffic to your site with ads. That's not necessarily very representative for the people that you would normally have read your resumés.

"versus", not "verses"

Reply on news.ycombinator.com