What is the ideal time fo the week to ask a Ask Mefi question to maximixe views and responses. May 20, 2009 7:16 PM Subscribe
So I've asked a few AskMefi questions now, and have noticed that they tend to get responses for about five hours, give or take, then the responses stop. When is the optimal time during the week to ask my AskMefi question, so that I get the maximum number of eyeballs looking at my question, and thus, get the maximum number of answers. I know that this may not necessarily increase the quality of the answers I get. I am talking about sheer quantity. Thanks!
Noon-5 PM Eastern US time, on a weekday.
1) Most users are American.
2) Noon-5 PM Eastern is 9-2 PM Pacific, both of which are decent windows.
3) People go out on the weekend.
Given that, I don't think it matters much. "If I just had more eyeballs on my question, I'd have better answers" is only true if the eyeballs you get aren't representative of the total eyeball population.
posted by smackfu at 7:23 PM on May 20, 2009
1) Most users are American.
2) Noon-5 PM Eastern is 9-2 PM Pacific, both of which are decent windows.
3) People go out on the weekend.
Given that, I don't think it matters much. "If I just had more eyeballs on my question, I'd have better answers" is only true if the eyeballs you get aren't representative of the total eyeball population.
posted by smackfu at 7:23 PM on May 20, 2009
Oh, and this has been covered a billion times before. Like this one about optimum posting time.
posted by smackfu at 7:28 PM on May 20, 2009 [1 favorite]
posted by smackfu at 7:28 PM on May 20, 2009 [1 favorite]
Repeating your question almost verbatim in the tags usually attracts more answers.
posted by gman at 7:29 PM on May 20, 2009 [2 favorites]
posted by gman at 7:29 PM on May 20, 2009 [2 favorites]
More important than time is to make your question jump out at people, so they want to read it and then answer it. It doesn't matter what time you ask "Third party shipping insurance - any recommendations?" ain't nobody going to read it or answer it, cause its boring! You got to blazz it up a little. Try "I recently slept with my blind brother's wife and read her e-mails without her permission and found out that she prefers black guys. How do all black people feel about keyloggers and is there a third party shipping insurance I can use when I mail my brother a tape I secretly recorded of me and his wife having surprise anal sex?" Doesn't matter what time you ask that bad boy, you will get results!
posted by ND¢ at 7:36 PM on May 20, 2009 [19 favorites]
posted by ND¢ at 7:36 PM on May 20, 2009 [19 favorites]
The problem is, when there are more people reading and potentially answering, there are also more people composing questions after you did, to push your question off the front page. I think it's pretty much a wash.
posted by Chocolate Pickle at 7:47 PM on May 20, 2009 [2 favorites]
posted by Chocolate Pickle at 7:47 PM on May 20, 2009 [2 favorites]
More important than time is to make your question jump out at people, so they want to read it and then answer it. It doesn't matter what time you ask "Third party shipping insurance - any recommendations?" ain't nobody going to read it or answer it, cause its boring! You got to blazz it up a little. Try "I recently slept with my blind brother's wife and read her e-mails without her permission and found out that she prefers black guys. How do all black people feel about keyloggers and is there a third party shipping insurance I can use when I mail my brother a tape I secretly recorded of me and his wife having surprise anal sex?" Doesn't matter what time you ask that bad boy, you will get results!
Me, I prefer to go for wehter I should watch subbed or dubbed anime while delawing my circumcised cat.
posted by ShawnStruck at 7:48 PM on May 20, 2009
Me, I prefer to go for wehter I should watch subbed or dubbed anime while delawing my circumcised cat.
posted by ShawnStruck at 7:48 PM on May 20, 2009
Ironically, it was when you posted to MetaTalk instead.
But thanks for playing!
posted by yhbc at 7:50 PM on May 20, 2009 [1 favorite]
But thanks for playing!
posted by yhbc at 7:50 PM on May 20, 2009 [1 favorite]
The key in my opinion is to have enough information on the outside so that someone who can answer your question KNOWS they can answer your question.
"I just found out I need major surgery, and I've got some questions... " (saying what the surgery is will get better answers)
"How far do you go before you call it quits?" (WAY too vague)
"How do you balance employee rights vs company needs in an employment contract?" (sounds theoretical but it's really a specific example for software developers)
"How to say goodbye to someone you love? " (vague)
"Alternative-biochemistry-filter " (not a question)
Main problem is that people are way too terse on the main page. Especially with relationship questions.
posted by smackfu at 7:51 PM on May 20, 2009 [1 favorite]
"I just found out I need major surgery, and I've got some questions... " (saying what the surgery is will get better answers)
"How far do you go before you call it quits?" (WAY too vague)
"How do you balance employee rights vs company needs in an employment contract?" (sounds theoretical but it's really a specific example for software developers)
"How to say goodbye to someone you love? " (vague)
"Alternative-biochemistry-filter " (not a question)
Main problem is that people are way too terse on the main page. Especially with relationship questions.
posted by smackfu at 7:51 PM on May 20, 2009 [1 favorite]
This question has indeed been asked a ton of times. It does strike me that it has an empirical answer, and maybe this would be a good project for one of the resident data crunchers.
posted by Miko at 8:13 PM on May 20, 2009
posted by Miko at 8:13 PM on May 20, 2009
Looking at your history, you really don't have a problem eliciting answers.
posted by gman at 8:21 PM on May 20, 2009
posted by gman at 8:21 PM on May 20, 2009
Data is so crunchy.
posted by IndigoRain at 8:25 PM on May 20, 2009
posted by IndigoRain at 8:25 PM on May 20, 2009
maybe this would be a good project for one of the resident data crunchers.
That gets said almost every time this comes up, but no one's bothered to crunch the numbers. Which probably means the question's not really that interesting. Anyway, different kinds of questions probably have different optimal times for posting, so it's not clear how you'd make the data meaningful.
posted by mediareport at 8:43 PM on May 20, 2009
That gets said almost every time this comes up, but no one's bothered to crunch the numbers. Which probably means the question's not really that interesting. Anyway, different kinds of questions probably have different optimal times for posting, so it's not clear how you'd make the data meaningful.
posted by mediareport at 8:43 PM on May 20, 2009
One problem is that while "when to ask to get the most answers" is answerable, it's not the real question. Rather, people want to know "when to ask to get the best answers", which requires some kind of shaky assessment of the quality of the answers.
posted by smackfu at 8:46 PM on May 20, 2009
posted by smackfu at 8:46 PM on May 20, 2009
1p - 3p edgeways standard time. All questions answered.
posted by edgeways at 8:57 PM on May 20, 2009
posted by edgeways at 8:57 PM on May 20, 2009
Plutor actually did crunch the numbers pretty recently, and generated a sort of scandalous result: it doesn't really matter. Noise dwarfs local variation, basically, so ask your question as well as you can and don't worry about the when.
Some more fun from that same thread.
posted by cortex (staff) at 9:06 PM on May 20, 2009 [2 favorites]
Some more fun from that same thread.
posted by cortex (staff) at 9:06 PM on May 20, 2009 [2 favorites]
I usually try to post my questions on days ending with a "y" between 12AM and 11:45 PM.
posted by not_on_display at 9:14 PM on May 20, 2009
posted by not_on_display at 9:14 PM on May 20, 2009
RIGHT NOW!
posted by abc123xyzinfinity at 9:33 PM on May 20, 2009
posted by abc123xyzinfinity at 9:33 PM on May 20, 2009
Tagging & categorising appropriately seem to be more useful than trying to get the timing just right.
posted by batmonkey at 9:35 PM on May 20, 2009
posted by batmonkey at 9:35 PM on May 20, 2009
What's totally awesome about this schema is that it's more or less self-defeating.
The more people that post in any prime time slot the less it stays on the front page and the less time each answerer has for answering various questions.
posted by loquacious at 9:47 PM on May 20, 2009
The more people that post in any prime time slot the less it stays on the front page and the less time each answerer has for answering various questions.
posted by loquacious at 9:47 PM on May 20, 2009
A half hour later in Newfoundland.
posted by Alvy Ampersand at 10:51 PM on May 20, 2009
posted by Alvy Ampersand at 10:51 PM on May 20, 2009
As loquacious says (and Plutor verified with data), for sheer number of eyeballs on your question it doesn't matter. But if you have a question that is best answered by a class of user that is more likely to be on at a particular time, try to ask at that time. More detail here.
posted by grouse at 11:40 PM on May 20, 2009 [1 favorite]
posted by grouse at 11:40 PM on May 20, 2009 [1 favorite]
From the MetaAnalysis page on the wiki, there's the Plutor graph(s) mentioned by cortex above (filed under "Analysis of Threads / Posts," though "By time of day" is probably better), & also a graph from Burger-Eating Invasion Monkey, placing the average number of responses at 13 to 15, roughly.
posted by Pronoiac at 11:50 PM on May 20, 2009
posted by Pronoiac at 11:50 PM on May 20, 2009
> That gets said almost every time this comes up, but no one's bothered to crunch the numbers. Which probably means the question's not really that interesting.
I'm a datacrunchy type, and my gut response to this question is that it doesn't matter when you ask it, so the data is moot. A couple of observations based on how I use the site, and i suspect most people do:
1. I only check at certain times. If it is a slow period (middle of the night, weekends), then there will only be a handful of new questions, and I will look at them more carefully, just because I have nothing better to do.
2. If it's during peak (weekday, workhours), I still only check at certain times, but when I do there are literally hundreds of new questions, and I am forced to scan the titles for interesting topics, because I don't have the time to read the entire internet daily.
3. In both of the cases above, I only click on questions that are interesting to me (either as someone interested in an answer or as someone who could answer it).
I think this is the more important factor. If you post at a slow time, all the people who are around at the active period will see your question anyway, since nothing new has happened since they last checked. But what drives people to open, read and respond has more to do with their interest in the subject. If you post a "help me identify this song" thread, there are people who love doing that shit and are walking encyclopedias of music that will click on it when they get around to seeing it. I personally click on computer related stuff, because I am a tech person and can answer most of tech questions competently.
I suspect you will get more actual responses if you post mon-fri between 12pm and 3pm (hours where both east and west coast people working office jobs will be sitting around wasting time on the tubes), but it's highly debateable whether they will be of higher quality.
BTW I would crunch said numbers just for the hell of it if I had access to the raw data.. does that exist?
posted by cj_ at 12:56 AM on May 21, 2009
I'm a datacrunchy type, and my gut response to this question is that it doesn't matter when you ask it, so the data is moot. A couple of observations based on how I use the site, and i suspect most people do:
1. I only check at certain times. If it is a slow period (middle of the night, weekends), then there will only be a handful of new questions, and I will look at them more carefully, just because I have nothing better to do.
2. If it's during peak (weekday, workhours), I still only check at certain times, but when I do there are literally hundreds of new questions, and I am forced to scan the titles for interesting topics, because I don't have the time to read the entire internet daily.
3. In both of the cases above, I only click on questions that are interesting to me (either as someone interested in an answer or as someone who could answer it).
I think this is the more important factor. If you post at a slow time, all the people who are around at the active period will see your question anyway, since nothing new has happened since they last checked. But what drives people to open, read and respond has more to do with their interest in the subject. If you post a "help me identify this song" thread, there are people who love doing that shit and are walking encyclopedias of music that will click on it when they get around to seeing it. I personally click on computer related stuff, because I am a tech person and can answer most of tech questions competently.
I suspect you will get more actual responses if you post mon-fri between 12pm and 3pm (hours where both east and west coast people working office jobs will be sitting around wasting time on the tubes), but it's highly debateable whether they will be of higher quality.
BTW I would crunch said numbers just for the hell of it if I had access to the raw data.. does that exist?
posted by cj_ at 12:56 AM on May 21, 2009
Er. The Burger-Eating Invasion Monkey graph.
cj_: You want the infodump: cortex generated it, & also posted it at a mirror when the original location had problems. I've posted some notes on working it over with Excel, if that helps.
posted by Pronoiac at 1:32 AM on May 21, 2009
cj_: You want the infodump: cortex generated it, & also posted it at a mirror when the original location had problems. I've posted some notes on working it over with Excel, if that helps.
posted by Pronoiac at 1:32 AM on May 21, 2009
Oops! I'm sorry this question has been asked a billion times before.
I was originally going to post it in AskMefi but then thought it would
have been moved here, so I came here, instead.
Everyone here is funnier.
Maybe it's the institution-gray background.
posted by Sully at 5:57 AM on May 21, 2009
I was originally going to post it in AskMefi but then thought it would
have been moved here, so I came here, instead.
Everyone here is funnier.
Maybe it's the institution-gray background.
posted by Sully at 5:57 AM on May 21, 2009
Oops! I'm sorry this question has been asked a billion times before.
Eh, it happens. Backtagging metatalk might help going forward, but there's a certain amount of repetition that's gonna happen around here with as many users as we have.
You were right to ask it here, regardless; if you'd posted it in AskMe we would have just nixed it and told you to bring it over here anyway. Questions about the site go in the grey, not the green.
posted by cortex (staff) at 6:34 AM on May 21, 2009 [1 favorite]
Eh, it happens. Backtagging metatalk might help going forward, but there's a certain amount of repetition that's gonna happen around here with as many users as we have.
You were right to ask it here, regardless; if you'd posted it in AskMe we would have just nixed it and told you to bring it over here anyway. Questions about the site go in the grey, not the green.
posted by cortex (staff) at 6:34 AM on May 21, 2009 [1 favorite]
Funnier, yes, but also meaner. MeTa is kind of like everyone's had four beers and are now eyeing number five.
posted by SpiffyRob at 6:37 AM on May 21, 2009 [4 favorites]
posted by SpiffyRob at 6:37 AM on May 21, 2009 [4 favorites]
> One problem is that while "when to ask to get the most answers" is answerable, it's not the real question. Rather, people want to know "when to ask to get the best answers", which requires some kind of shaky assessment of the quality of the answers.
Well, let's look at the question, shall we?
I know that this may not necessarily increase the quality of the answers I get. I am talking about sheer quantity. Thanks!
Gosh, that was easily refuted! Now, who's buying the next round?
posted by languagehat at 6:45 AM on May 21, 2009
Well, let's look at the question, shall we?
I know that this may not necessarily increase the quality of the answers I get. I am talking about sheer quantity. Thanks!
Gosh, that was easily refuted! Now, who's buying the next round?
posted by languagehat at 6:45 AM on May 21, 2009
A student once asked Master Zhaozhou, "Master, how long must I wait for the best time to post to AskMetafilter?" Zhaozhou replied, "I am busy now with my calligraphy. Come back in an hour." An hour later the student returned and repeated his question. And Zhaozhou replied, "No, I am still too busy. You must be patient. Come back in two hours." The student did as he was told, and after two hours returned to Zhaozhou and asked his question a third time. Zhaozhou replied, "But I have already answered your question." And the student was enlightened.
posted by chinston at 6:55 AM on May 21, 2009 [1 favorite]
posted by chinston at 6:55 AM on May 21, 2009 [1 favorite]
There are two good times to ask questions:
1. Between noon and midnight
2. Between midnight and noon.
Either of those options will generate answers.
posted by blue_beetle at 6:56 AM on May 21, 2009 [1 favorite]
1. Between noon and midnight
2. Between midnight and noon.
Either of those options will generate answers.
posted by blue_beetle at 6:56 AM on May 21, 2009 [1 favorite]
Gosh, that was easily refuted! Now, who's buying the next round?
You are right. They do want to know the best time to get more answers, regardless of quality.
If the questioner just wants more answers, and doesn't really care what they are, just ask a relationship question, any time of day. Or a music poll. Or read other people's questions and imagine their answers are on your question.
posted by smackfu at 7:56 AM on May 21, 2009 [1 favorite]
How many questions posted at the optimum time would it take to decrease the maximum amount of overall answers as a functional response to the handling time of each individual answer per question, is what I want to know.
posted by kuujjuarapik at 2:57 PM on May 21, 2009
posted by kuujjuarapik at 2:57 PM on May 21, 2009
First of all, not everything in your life needs to be gamed, or as the kids like to call it these days, "hacked".
Second, any empirically definitive statement about the "best time to post" will just result in a spike in users posting during that time, which will necessarily make it no longer the best time to post. Think of this as The Law of SchroedingerFilter.
posted by mkultra at 7:08 AM on May 22, 2009
Second, any empirically definitive statement about the "best time to post" will just result in a spike in users posting during that time, which will necessarily make it no longer the best time to post. Think of this as The Law of SchroedingerFilter.
posted by mkultra at 7:08 AM on May 22, 2009
> You want the infodump: cortex generated it, & also posted it at a mirror
What timezone are these stamps? UTC?
posted by cj_ at 2:27 PM on May 24, 2009
What timezone are these stamps? UTC?
posted by cj_ at 2:27 PM on May 24, 2009
I found something kind of interesting. If you calculate the average comments generated per post by hour, it's pretty flat:
I'm not sure what to make of this offhand without looking more closely, but it confirms my suspicion that it doesn't matter when you ask your question. FWIW, ~9pm Pacific seems to generate more comments-per-question, but not by anything statistically signficant. The average comments per post is almost 14 on the dot. For the main site, it's a little noisier but still pretty flat at around 31 comments-per-post.
When I get more time I'll see if there's any trends with favorites/unanswered questions, etc., and see if day-of-week makes any difference.
posted by cj_ at 6:05 PM on May 24, 2009
hour average comments per post
0 13.87
1 13.01
2 13.04
3 13.49
4 13.79
5 14.00
6 14.08
7 14.15
8 13.93
9 13.77
10 14.05
11 13.96
12 13.88
13 13.89
14 13.80
15 13.72
16 13.78
17 14.23
18 14.45
19 14.29
20 14.10
21 14.66
22 14.46
23 14.47
I'm not sure what to make of this offhand without looking more closely, but it confirms my suspicion that it doesn't matter when you ask your question. FWIW, ~9pm Pacific seems to generate more comments-per-question, but not by anything statistically signficant. The average comments per post is almost 14 on the dot. For the main site, it's a little noisier but still pretty flat at around 31 comments-per-post.
When I get more time I'll see if there's any trends with favorites/unanswered questions, etc., and see if day-of-week makes any difference.
posted by cj_ at 6:05 PM on May 24, 2009
You are not logged in, either login or create an account to post comments
posted by klangklangston at 7:18 PM on May 20, 2009