Link Buying, Anchor Grams, UGC, Persuasive Influences and Linking Oot n Aboot – Scratchpad

Today’s scratchpad is going to cover some new how to ideas in link building, the ROI on Google’s user generated content (UGC) properties, influence and then do some linking oot n aboot. If you enjoy this, please subscribe.

Link Buying and Building – Here are a couple of tips to get your anchor text profile looking more natural as well as find sources for buying links.

  • 5ubliminal at TellinYa opened my eyes to N Grams. N grams are a math idea that essentially looks at patterns in what follows and/or precedes a set item. So when the items are words, n grams consider the words following/preceding a given [key]word.
    • Turns out that Google’s been using them for a while. (Hat tip Quadszilla) I daresay that they’re not so effective at using it yet, or else they don’t care that much when they know that all the players in an industry have their hands dirty. But as in chess, you can’t count on your opponent (here, Google), not making the best move. And integrating N grams is really a great move to make.
    • Do you see where this is going yet?
    • Read a few sources before picking what your anchor text is going to be for a given link building campaign. Then, when you know what common preceding/following words are, you can mix those into your set of targeted anchor text.
    • If you care enough, think this is a valuable enough tool, you can buy Google’s released data on n grams. (Feel free to buy me a copy too ;D). Beware though that there are some doosies in there, so you’ll want to think twice if some given n gram doesn’t make sense to you. Google’s seeded that with ‘poison keywords’ (idea from Mike, confirmed by Matt; again, sourced from Quadszilla).
    • If you want to do this for linkbait, you can scrape together a massive amount of text a la Google and then provide an n gram generation tool for folks to figure out the n grams for their keywords. Or you can base the tool on Google’s larger data set, and just share the caveat that some of the results will be nonsense.
  • Link buying – I’ve IDed a footprint. Search around for sites that, when linking without nofollow, post stuff like ‘full disclosure, these guys are advertisers of ours’. You can buy on that same site, or get your purchases from other sites those people own. That said, Google’s probably IDed that same footprint …

What’s the ROI on user-generated content? The following’s some commentary and questions in note form, so brace yourselves.

  1. Blogspot – Hosts plenty of spam, captcha is broken. Spam flagging just triggers on initial affiliate monetization/adsense use (kind of ironic considering that they encourage adsense’s use almost immediately). The ROI here is debatable.
  2. Jotspot – Funny, but I don’t hear anything on the proliferation of Jotspot wikis. Adsense integration must suck. UGC means typos OK; spam by scrape and replace.
  3. Blogsearch – Feeds traffic to blogspot spam. Blogsearch should be available off main engines’ home pages, just by clicking a radio button (better than tab links). Is the index the whole web, or just blogs? SERPs by recency suck. See also my analysis of Google’s Blogsearch algorithm.
  4. I should have something to say on Youtube, but suffice it to say that at $15M annual revenues, it’ll take a good while before Youtube pays for itself. Also, I think this purchase was the official trigger that inflated the Web 2.0 bubble. Edgeio (classifieds by TechCrunch) was the first to die, I just got an email about some feed analytics provider dying… More will be coming soon, I’m sure.

Motivation tied to Cialdini’s 6 principles gets you influence. Maki explained Cialdini’s 6 principles, and I recommend the read. Now, I’ve got some more to share on things you can do to influence folks.

  • Time magazine was recently presenting some scientific research on monkeys’ social behaviour. The monkeys had to eat all together. If some were missing, the whole group went hungry. One night, a couple of monkeys decided to monkey around and go MIA. The other group had to wait a long time before the keepers found them and brought them to the eating area. Anyways, the group who showed up for their food on time got really annoyed and beat the two miscreants the next day (the keepers kept them separate from the main group overnight because they could tell that the main group was agitated). Speaks volumes about the power of reciprocity, doesn’t it?
  • Social proof: Using this can increase your credibility (contrast that with these major credibility mistakes). So next time you’re designing a site or upgrading, consider showing off these:
    • Sponsorships of events/charities etc. This proves size. Some people are reluctant to deal with smaller companies and this could help them classify you in the big-and-reliable category (in their mind anyways; I think there are plenty of problems with big companies, like 1and1 and Network Solutions).
    • Have a degree? Make it known. Share your M.Sc in the site description or post a picture.
    • Write a book. Note: ebooks don’t give a fraction of the credibility writing a real printed book does. Ditto if the book is self-published versus published by a large publisher.
    • Let people know that you’ve been featured in mass media. Rich Schefren is a spammer, but if you look at his site, he’s apparently been all over the major news outlets. Note: None of those are linked to the actual stories, so it may be a crock too; linking to the sources would be a good idea.
    • Be a regular columnist/correspondent in the mass media. As awful as the mass media can often be, biased and erroneous, they are followed by millions and largely trusted as being authoritative.
    • Write guest articles. Yours truly has done this for Sitepoint, BrandCurve, Pandia, ISEdb, Huomah, SEOmoz etc.

Finally, check out these links you’ll love.

Speaking of n grams, how do you like this mega resource of free books? The good kind, not the clickbank get-rich-quick trash kind.

Handsome Rob’s got a great entry on why most newbies fail with affiliate marketing.

Annie’s got the recipe to make your blog flavourful.

Jane’s pointing out (like Ciaran before her) that your offline ad campaign needs to work with your SEO campaign.

How would you like a bulk domain appraisal tool?

Brian Chappel’s got 30 link buying best practices (which I augmented by 2 in the comments).

Care about your feed subscribers? See how Trent went 0 – 12K in no time. Then check out my related material on metrics to measure your success with social media. Which, if you subscribe, you’d have already noticed…

Shady’s got some points on how spammy promotion won’t necessarily hurt your brand. On a related note, if autogenned spam earns affiliate commissions, is it really all that bad? I mean, the visitor got what they wanted, and isn’t that the best metric you can use… I know Shady does well with his affiliate oriented spamming, which is why I ask.
Last but not least (best, in fact) – see Scratchpad numero uno. Ok maybe not the best, but I’ll bet you like it. And you’ll definitely want to subscribe to my RSS feed by the end of it :D. (You can always click the feed link in the blue box on the upper left if you navigate away from this page and don’t care to return here just to find the subscription link.)

Tags: , , , ,

Comments

  1. The N-Grams are cool to know about as a concept and for self-use. Google's data is useless to the average blackhat. First thing is it's so large that handling it on a PC is ... unlikely to happen. On the other hand you can just scrape your niche and build your own N-Grams. This will make them topic related and much better then hugenormous data Google provides. What Google actually means by tellin'ya about N-Grams is: We know more then you would imagine! We can slap you anytime. The problem is they care about index size and too many dirty players out there to move on them too hard. I'd say 75% of index could be easily labeled as cr4p. Don't think of buying Google's data:) The chances to find use for it are so low unless you got some datacenters of your own. Regards.

    Comment by 5ubliminal - February 10, 2008 @ 7:44am
  2. Wow, that was pretty insightful 5ub! I didn't realize that the data was too large to manipulate on a PC. Sounds like trying to herd elephants along with a twig. Glad to hear about Google telling people they can be slapped, too. That's always nice :). I'm not sure about the 75% estimate (unless you've got some research of your own to back it up?) but certainly there's a load of crap out there, as anyone who's done link research on their competitors can attest.

    Comment by Gabriel Goldenberg - February 10, 2008 @ 5:23pm

Leave a Reply