Well, I got over 4,000 emails responding to my research assistant gig posting. It genuinely flabbergasted me, especially after I took pains to explain to people that this was not a job that involved any sort of goofy partying with me or whatever, and though it pays, it isn’t much at all. I was hoping to get 500 responses.
Of over 4,000 emails sent to me, only about 1,500 got replies back inviting them to the second round. Yes, you read that correctly: Only 37.5% of the people who responded could follow the extremely basic instructions I gave.
Since I have closed the first round and the second round has started, I’m going to go through each instruction and explain how people got them wrong, because I think it might be beneficial to people who are in the second round to understand my thinking. It also might help some of those people who fucked up to understand what they look like to an objective employment source, and learn something from their mistakes.
Here are the full instructions I put up:
1. Before 5pm EST on January 24th, send an email to this email address: firstname.lastname@example.org
2. The subject line must read “Tucker Max Research Assistant Gig”
3. In the email, put four things: 1. your full name, 2. how old you are, 3. where you live, and 4. what the third largest magazine in the US is, by circulation
4. Attach a picture of Jason Williams to the email (not the basketball player). It must be either a gif or a jpg, and cannot be larger than 1 megabyte (1MB).
5. Do ONLY those tasks. Nothing else should be in the email.
1. No one got the due date wrong, because I had to cut off the applications way early because of the overwhelming volume. This was not something I planned on doing, and if you waited to submit your email for some reason, I’m sorry. But considering it would only take about five minutes to accomplish the tasks I set out, I don’t know why you would wait.
The funny thing is, five people sent emails to my personal email address. Which is not listed anywhere on this application (it is on my other website, TuckerMax.com).
2. This instruction seems pretty obvious right? I accepted “Tucker Max Research Assistant Gig” with or without quotations marks, and still I was able to eliminate 1% of the applicants based on getting just this simple instruction wrong.
3. This question was the meat of it. No one could really get 1., 2. or 3. “wrong” per se. But a ton of people to put in forgot various elements, and were immediately eliminated.
The big thing was part 4. “what the third largest magazine in the US is, by circulation.” This is not hard to find information. I mean, it’s just fucking listed on Wikipedia, so it can’t take more than two seconds, right?
Did you really think I want to hire someone who would just go to Wikipedia and uncritically accept whatever was there? No fucking chance. I want to hire a research assistant who can actually think a little bit, someone who understands that Wikipedia is complete shit as a sole source and must ALWAYS be checked against other sources.
So before I posted this gig, I went to Wikipedia and I switched the #3 and #5 slots. Unfair? Maybe. But did it accomplish my goal of weeding out people who can’t think critically? Even more than I thought it would.
You see, almost as soon as I posted this job, other people went to Wikipedia and fucked with the #3 result as well. Whereas I had put in a magazine that could be reasonably thought to be #3, people started vandalizing Wikipedia with ridiculous shit. And the best part: Other people just repeated it back to me! Think I’m joking? Here’s what HUNDREDS of people put as the answer to #4:
I Hope They Serve Beer In Hell (over 120)
Fuck Off And Die (over 20)
Swallow or sleep in the wet spot (over 160!)
Assholes Finish First (over 100)
Wall Street Journal Magazine
I can track the changes exactly in the emails I get, because at least half of the people just mindlessly copied what was in there. I’d get 100 “I Hope They Serve Beer In Hell” and then it would switch and never go back, and I’d get 140 of the next, and so on. I was flabbergasted by this. I mean, I could see someone just accepting a false entry that was actually a magazine without checking other sources, but who would think “Swallow or sleep in the wet spot” was a fucking magazine?? Apparently, about 160 people did.
Some people did the smart thing, and accounted for it, I guess because they thought I might be testing them in some weird way. They wrote that what the obviously wrong answer on Wikipedia was, and then listed another source with the real answer. Several people did that, and they advanced to second round.
In case you are wondering, there were two potential correct answers: Costco Connection is the largest in terms of total circulation, and Better Homes and Gardens is the largest in terms of paid circulation. I accepted either. The best answers had both and explained the difference, but there were only like 10 of those. I expect those people to do well in the round 2. A lot of people thought Readers Digest was the right answer. It would have been, in 2009. I wasn’t clear about the year in the question, but it should be obvious that I’m talking about currently. There were two people who mentioned this and put the right answers for various years, and I allowed that of course. More research is always good, if its correct.
I actually fucked up in my question here, and could have eliminated a ton more people. I meant to have the word “paid” before “circulation” in the question. This is because the top two entries for the question “what the third largest magazine in the US is, by circulation” are Wikipedia, which I altered, and a source called NyJobsSource, that reprints the National Magazine Councils circulation numbers. The National Magazine Council counts paid and unpaid periodicals, whereas to get the “paid circulation” numbers you have to dig a little on Google, and actually read some to understand what different agencies actually measure. I am after all hiring a “research assistant” not someone who spits Google or Wikipedia results back at me, and real research is more than using a search engine. It means understand not only the nature of the question, but the source the data is coming from, and what the data actually says.
But even though I didn’t do that this task eliminated by far the most people; I’d say about 1800 of the 4000 got this wrong.
And there were a lot of people, I’d say maybe 75, who just repeated the questions in the email back to me, like I was trying to trick them. Their answer was literally, “1. your full name, 2. how old you are, 3. where you live, and 4. what the third largest magazine in the US is, by circulation”.
Pretty much everyone who did this was between the ages of 16 and 20. This is some retarded fantasy fiction shit answer. Look children, I’m hiring someone to do an actual job by testing a basic skill (following instructions), not looking for an elf to solve the riddle of the Sphinx. That sort of answer works in video games, not real life.
4. The forth task was “Attach a picture of Jason Williams to the email (not the basketball player). It must be either a gif or a jpg, and cannot be larger than 1 megabyte (1MB).”
I have to apologize to my friends here–everyone I know got emails from random people asking which Jason Williams I was looking for. It didn’t matter; it could be one of 10 different ones; there’s a football player, a hockey player, an actor, etc. The point was to test three things: 1. Can someone follow simple instructions about attaching a picture and size/format, 2. What will they do when operating under a situation of sub-optimal information, and 3. What was their level of cultural engagement.
If you Google “Jason Williams pics” most of the first pictures are of the white Jason Williams basketball player, and the two black Jason Williams basketball players. In many of them, they are playing basketball, but many they are not. I’d say at least 300 people used pictures of a basketball playing Jason Williams, but out of uniform. Again, you are applying to be a RESEARCHER, you need to be able to think, not just spit Google results back at me.
You may not be shocked by this, but at least 100 people put not just a picture of one of the three basketball playing Jason Williams, but they put picture of THEM PLAYING BASKETBALL!!! Like, with a basketball in their hand. WTF?
5. I thought this would catch more people than it did, “Do ONLY those tasks. Nothing else should be in the email.” I explicitly said that this was a measure of how well you followed instructions, but I figured a bunch of people would add their resume or try and tell me how much they wanted to work for me. Very few did that, maybe 50.
One thing I should say: Sending out 1500 emails is a fucking pain in the ass. I had to use all the Gmail keyboard shortcuts to make it faster, and it still took forever. But the problem with using keyboard shortcuts that is that sometimes I accidentally fucked up. There were two fuck-ups for me:
1. Sending an email to someone who got their email wrong: I did this I think about 15-20 times. For example, someone would have a picture of the Duke basketball player Jason Williams in a suit, but get everything else right, and I would accidentally hit send before I realized it. Which means there are some of you in round 2 who don’t deserve to be. Oh well, mistakes happen, and you got lucky.
2. Not sending an email to someone who got everything right: I was generally OK with making the top mistake; if the person is an idiot, I’ll definitely figure it out in round 2, no harm done. I didn’t want to make a mistake where someone got everything right and I skipped their email, so I tried to er on the side of sending versus non-sending. I don’t *think* I skipped anyone who got all the instructions right, but if you sent me a correct email, and did NOT get a response back, do this: Send me an email, to the address you sent the original email, with the subject line “You made a mistake.” I will then search my inbox for your old email (though Gmail should thread it anyway). If you’re right, I’ll send you the second round instructions. If you’re not, I’ll call you an idiot.