|
May 29
2008
|
Do you spend your days in the drudgery of promoting your web sites? We've listed over 2,000 Social Media Websites as ripe grounds for promoting your articles. But the comment that we seemed to get the most was best summed up by katfrench on Sphinn:
"Seriously, though, post a list of two thousand social bookmarking sites and Sphinners will get a headache at the thought of all the time it would take to register.""
Why You Shouldn't Try to Submit to 2,000 Social Networks
- You Shouldn't Submit Your Own Stuff. I agree that there's nothing morally wrong with promoting yourself, but most of the communities out there will tend to vote you down if you just blow your own horn. It is far better to have a variety of people submitting your articles to the social networks, if only because then you've got a wide range of IP addresses submitting your articles. And let's not forget that Google was making noises earlier this year about devaluing links from social networks where it appears that the owner of the site is linking to their own pages.
-
It's a poor use of your time.
How much are you willing to pay for links? How much is your time worth? It's a matter of looking at your investment of time and how much a link or traffic is worth to you. If you decide a link from a third tier social network is only worth about a nickel and it takes one minute to submit the article, you're saying your labor is only worth $3/hour to do this yourself. If you live in the US, you can't live on that.
Why You Should be on as Many Social Networks as Possible
This sounds like a contradiction with the first section, but it really isn't. There are a lot of very good reasons that you should be on as many social networks as possible, instead of just focussing on the first tier like Digg, Reddit, Del.icio.us, etc.
- Cover Your Bets. What if you had spent hundreds of hours building up your site on Digg and they decided to ban you? Or what if Google's solution to the social nework self linking problem is to just dramatically discount all links from Digg? Concentrating on just a few social networks is really putting all of your eggs in one basket. It's dangerous and foolhardy. You should spread your risk across as many platforms as you can.
- Little Guys Bring Better Traffic. Yes, StumbleUpon and Digg can bring massive traffic, but the bounce rate is also astromical. Is it really worth paying for the bandwidth to bring in 100K visitors that spend an average of 15 seconds on your site? Who aren't really interested in what you have to sell? On the other hand, getting 5-10 good targeted visitors from a specialty site can lead to a conversion rate that will make it worth your while. It's just a question of whether you want to be able to brag about how much traffic you have, or whether you want to make money because you've got real visitors that are interested in what you have to sell or will click your ads.
- Little Guys Have Less Distance to the Top. Your submission on Digg will end up several levels away from the top page with the big PR boost. Even if it makes it hot, it will fall off and the long term link juice gained will be fairly minor. But on a small social site, your submissions will be fairly near the top. Would you rather be 10 steps away from a PR 9 page, or two steps from a PR 4? My gut tells me hitting singles all the time is better than hitting home runs but batting .050.
- Little Guys Aren't Filled With Jerks. There's no "bury brigade" on the small sites. The editors of the sites are pleased to have people submitting stories, as a lot of them end up seeding their own stories
- The Link Math is Astonishing If you concentrate on the top 3, you can at most get away with submitting 3 articles a day across them. And you'll have to bury that within many others each day so you don't look like you're only self promoting. 3/day = 15/week = 750/year. So for a year's worth of work you'll get 750 links, and if you're lucky they'll have an average PR of 1. That's not going to do much for you. Now what if you targeted 400 sites? The math now works out to 400/day = 2,000/week = 100,000/year. Now we're talking.
So Build a Bot, Right?
Wrong! Even if you could code a bot that would handle submissions on 400 different sites, it just wouldn't work.
- It's from the same IP address. Editors tend not to like bots (imagine that!), and your bot is coming from either a single or a small group of IP addresses if you use a proxy. Banning your IP is pretty darn easy. That's a big investment in development that you would make and it would be gone.
- It's from the same people. Unless you also maintained a huge stable of fake accounts on each of these sites (which would also have the same IP problem), your submissions would all be coming from the same people and could be easily banned.
- These sites change. There are several "auto-posters" that have been defeated merely by adding a little bit of javascript to the login. The bot's Curl script can't deal with that, so they cough. Just moving things around the page or changing the names of input fields is enough to make maintenance of your huge catalog of bot scripts impossible. You'd need a full time staff just to keep up with it.
- Captchas. Most of these sites are moving to some sort of Captcha on their submission, so unless you're an elite hacker it's not going to work. If Captcha's aren't a problem for you, then you don't need to be reading this.
- It Would Violate the TOS. Not that the bot spammers care, but most of these sites say you can't use a bot on them.
So What's the Solution?
The solution is the same one that worked the first time someone started a business: leverage the work of other people. All you have to do is get other people to post your stuff.
How to get them to do that and still maintain profitability is a lot less obvious. But you'll have to wait for tomorrow's post to find that out.

written by Ted Sisul, August 21, 2008
You talk about leveraging the work of other people to get your website posted onto Social Media websites; however, you mentioned explaining how and I didn't see an answer in your next post.
I thought about using Amazon's Mechanical Turk and creating Hits (virtually paying others to bookmark my website). Have you tried this? And if so, have you had any success? Also, do you have any other suggestions?
Thanks by the way for the great article. It shed a lot of light on questions I had regarding promotion vs. spam.
written by Don Draper, August 21, 2008
As for the Amazon Mechanical Turk, that's what I was alluding to. We experimented with the Turk system and had good success. But we found that the Amazon interface made creating and servicing HITs rather difficult for this type of thing. Expect us to release a product soon that simplifies and automates all that.
There's a lot more you can do with this approach than just paying people to bookmark your posts.
written by Ted Sisul, August 21, 2008
Thanks for the lightning fast response. It's much appreciated.
That's awesome! I can't wait for the release.
Also, I have a few quick questions on using Mechanical Turk until your software is available:
1. How much did you pay per HIT (one bookmark on one Social Media website)? I'm not looking for an exact number, but was it dollars or pennies? As well, did you find return performance was poor if you paid too little?
2. Would you recommend breaking out HITS as one event. Such as one bookmark for one page on say Kaboodle?
3. How many bookmarks (or HITS) would you recommend limiting to one user, so that you don't have the same problem with too many coming from the same IP?
Again, I really appreciate the help and look forward to the release.
written by Don Draper, August 21, 2008
People are always surprised when we respond quickly. I don't understand why so many companies work so hard to generate interest, then don't engage people that want to communicate.
1. How much did you pay per HIT (one bookmark on one Social Media website)? I'm not looking for an exact number, but was it dollars or pennies? As well, did you find return performance was poor if you paid too little?
Actually, it was the opposite. We offered HITs at 0.10, 0.05, 0.02, and 0.01. Same task, just different prices. We got the best response out of 0.02. My guess is that lots of turks keep checking for tasks in the 0.02 range because they're usually quick to do. So overpaying actually hurts you.
2. Would you recommend breaking out HITS as one event. Such as one bookmark for one page on say Kaboodle?
Yes. One HIT per task. Otherwise the verification becomes a real pain. We found that it was very important to have an auto-verification set up so that the turks see feedback right away that they're going to get paid. A lot of providers tend not to pay until the task times out on them, usually 10 days or so. Most turks will do a few tasks for you and then see if you pay in order to limit their exposure.
3. How many bookmarks (or HITS) would you recommend limiting to one user, so that you don't have the same problem with too many coming from the same IP?
If you do 1 HIT/task you don't have that problem with entering a story in a news site. If on the other hand you want 100 people to bookmark you in Technorati, make 1 HIT with a limit of 100 tasks. AMT will force it to be 100 unique people.
Ok, a few questions for you:
1) How much would it be worth per month to have access to a system that would let you enter a list of URLs to bookmark, select from hundreds of bookmarking sites, and push a button to generate HITs that hook into an auto-verification script that auto-pays the turk when they complete the task?
2) Is this something you'd use all the time, or just in projects of short duration?
written by Ted Sisul, August 22, 2008
Does the software also have the ability to distribute tasks using a schedule? I know from experience that you can trigger Google filters when obtaining a ton of links too soon, particularly when they have similar anchor text. It would be nice to load up all of your tasks and then have the software send HIT requests over a scheduled time frame. Something similar perhaps to what RSS Bookmarker claims it does.
As for your second question, if every thing worked like a well-oiled machine. In other words, tasks were sent out and completed swiftly and accurately, I would use a tool like this daily. The time saving and monetary potential would be unbelievable.
I have to say again, thanks a ton for your helpful advice, and please keep me posted on the software you're creating. I'm definitely interested.
written by Ted Sisul, August 27, 2008
I hate to bug you with another question, but I've starting using Mechanical Turk and have an issue I can't find the answer to. I emailed Amazon 4 days ago, but still haven't received a response. I also tried searching on Google for 2-1/2 hours today, but nothing relevant has come up.
You mentioned creating one HIT with a limit of 100 tasks/assignments, which AMT forces to be 100 unique workers. Can you tell me how I would create multiple assignments in one task? I've played around with the system for days and still can't figure this out. Say I put 100 URLs within one HIT (100 URLs to be bookmarked by 100 different workers on one Social Media website). How am I able to limit one Worker per URL, so that each bookmark comes from a different IP address?
Again, any help you could provide would be greatly appreciated.




cheers
daryl.mu