How many hits does a front-page MeFi link give the linked site?

April 16, 2004 9:39 AM

How many licks does it take to get to the center of a Toosie Roll Pop? How many chicks can a woodchick chuck? How many hits does a front-page MeFi link give the linked site? More inside...
posted by Shane to MetaFilter-related at 9:39 AM (16 comments total)

Okay, nevermind the first two questions. I'll save them for AskMe ;-)

A while back I posted something on the front page, and the site owner (whom I had never met) e-mail-thanked me later. Said he got an extra 500 hits. Now, it was posted at an off-peak time, and it may or may not have been an attention-grabbing post. But it was well-received, and the post certainly didn't, you know, totally suck. I'd cite the post, but I don't want to get into all kinds of particulars.

In general, it just made me think, because time was it seemed that ANYTHING posted to the front page brought anywhere from 1,000 to 5k hits. I have had feedback telling me a post brought 3,000 hits and another post 5,000 hits in 36 hours, but both of these were at least a year ago.

Am I wrong? Is the volume of posts on the front page lessening the hits a single post gets? What are your thoughts? I'm not criticizing, just curious.

Also, who knows?--the site owner's hit meter might suck. I installed several on my own site a while back, did a comparison, and found that the Site Meter whose configuration I like best only records 30% to 50% of my hits.
posted by Shane at 9:39 AM on April 16, 2004

Oh, I forgot to tell you the hilarious (and slightly joking) comment that will make you shudder:
[paraphrasing]
"I've been MeFied, but what do I have to do now to get Slashdotted?"
[/paraphrasing]
posted by Shane at 9:43 AM on April 16, 2004

I posted about this issue last August in connection with a thread about the weblog of The Homeless Guy, who reported receiving 900 hits in one day to his site due to a MeFi thread about his blog, with hits increasing for maybe a week or two afterwards, then dropping off in sad Flowers for Algernon manner. It can be hard to tell, though, because link may also have been seen at MeFi and then posted elsewhere -- difficult to attribute which hits sprang directly from MeFi.
posted by onlyconnect at 9:53 AM on April 16, 2004

I'm sure it'll vary depending on the flavor of the post, but, FWIW, I noted here that the egg-art site rec'd about 1,000 hits in a 20 hour period.
posted by Ufez Jones at 10:10 AM on April 16, 2004

Here's how the Homeless Guy described the hits he got from a MeFi thread about his site:

For the past couple weeks, since I started this blog, I've been averaging about 60 visits a day.

This past Saturday - 650
Yesterday, Sunday - 900
As of 9:30am Today -1100

I am actually trembling as I sit here, entering this blog. It's all overwhelming.

Later on, he described it this way: This posting on Metafilter generated some 900 visits to my blog that day. The following day an online magazine in the UK mentioned my blog (apparently after someone read about it on Metafilter) and my site received 1200 visits. The following day brought 3200 visits. The next day, Yahoo listed my blog on its list of "New and Notable" web sites, and it received over 15,000 visits.

I don't know whether the effect of a MeFi thread has dropped off since August of last year, but it's no wonder that someone like the ChristFilter guy would get an account just to make marketing posts, when a thread could give that sort of exposure. Always depends on the quality of the link, though (on preview, as Ufez says).
posted by onlyconnect at 10:14 AM on April 16, 2004

When the traffic from MetaFilter takes down servers, then you'll know it's become a force to be reckoned with.
posted by ralawrence at 10:27 AM on April 16, 2004

Are you saying that people actually read the linked sites? I may have to try that sometime.
posted by Kwantsar at 10:33 AM on April 16, 2004

When the traffic from MetaFilter takes down servers, then you'll know it's become a force to be reckoned with.

Good point. Similarly: does MeFi still kill GeoCities cites just by breathing on them? Just a brief mention on the front page used to overload their traffic limits. Um, not that MeFi has really bad breath. Or does it?

Are you saying that people actually read the linked sites? I may have to try that sometime.

That's a pretty astute point too, at least when applied to MeFi members...
posted by Shane at 10:38 AM on April 16, 2004

My site was FPP'd a good while ago (a year and a bit?). I was averaging about 100 hits a day prior to the FPP (there was nothing on my site but a subscription box so I come up in no searches unless searching for my name). After the FPP I got about 12k visits over the next three days and, if I remember correclty, about 180 subscribers. About two months later my site got nearly a full page write-up in one of Canada's large daily newspapers. In contrast, I recieved about 80 more hits than normal the day the article came out and a whopping 11 subscriptions. (Mind you, the newspaper article was anti-me.)

I'm sure the # of hits one gets varies drastically based on the topic (obviously) but also on the poster. I'm much more likely to click links by someone who's got a rep of good links (even if the post doesn't sound interesting) than people who I've learned to distrust. There are a few members whose links I never click, regardless of the topic. I would guess a lot of members make their decisions this way.
posted by dobbs at 11:16 AM on April 16, 2004

does MeFi still kill GeoCities cites just by breathing on them? Just a brief mention on the front page used to overload their traffic limits. Um, not that MeFi has really bad breath. Or does it?

no. anything looking crosseyed, from a distance, in a blizzard, kills a free geocities (apparently hourly) bandwidth quotas. i've consistently knocked them over even by viewing just a few times.
posted by quonsar at 1:54 PM on April 16, 2004

It's probably also worth keeping in mind the difference among "visits," "page views," and "accesses" -- any one of which could be called a "hit." One "visit" (extended session from one IP address) may comprise several page views. Each page view may involve several "accesses" (requests for files -- the HTML file, the image files, embedded files, etc.).

That is, 500 visits could easily account for 1500 page views, which could easily account for 6000 accesses. So a disparity in numbers may just reflect a disparity in accounting methods.
posted by macrone at 4:08 PM on April 16, 2004

i had a site on MF and then Slashdot two weeks later. The MF visits (unique IP visits) totaled around 3000 over two days but a lot of this might have been from other places too (but MF was the first to 'discover' it.) Then two weeks later Slashdot got it and I got 10x the visits in a couple of hours (and it was a late friday night, too. Shoudln't they have been watching "Family Matters"?)
posted by neustile at 4:28 PM on April 16, 2004

The thing about a MeFi post isn't so much the MeFi traffic, but that it's a hellmouth for link-spreading. A good link on MeFi will be picked up for weeks, and spread outward like a nuclear blast.

I always say this, because it's worth noting, but three years ago the most traffic I ever got on a site of mine was from a mention by David Pogue on the Martha Stewart Living television show. It was heavier than a Slashdotting at the time. I've since had a Slashdotting send about 65,000 visitors (not hits, mind you, but visitors) to a site in a single day. (I've seen three major Slashdottings and two minor, and every year they get harder to handle. It's not the throughput, it's the getting hammered by HTTP requests, which act like a DoS).
posted by Mo Nickels at 5:03 AM on April 17, 2004

HTTP requests, which act like a DoS

why is the usa patriot act silent on such matters?
posted by quonsar at 5:25 AM on April 17, 2004

why is the usa patriot act silent on such matters?

Wait for it.
posted by Jairus at 9:37 AM on April 17, 2004

Metafilter is discussed as causing the biggest spike in the memespread project over at kottke.com. According to the final writeup [pdf] on the project, "[w]hat seemed most interesting is that the largest spike occurred 10 hours into the experiment, and was due to the website MetaFilter (http://www.metafilter.com). The spread of the meme to this collaborative blog seemed to help give the epidemic another wave of spreading . . ."
posted by onlyconnect at 11:30 AM on April 30, 2004