For those of you who have not heard, widgetbait is dead. At least in it’s previous, useful fashion. To summarize, JustSayHi got the smackdown for distributing widgets that contained a link back to their online dating site. The technique worked incredibly well, ranking them for tricky keywords like “online dating” or “free online dating”. And them Google gave them the smackdown. So we’ve established that Google is not going to be happy if you actually rank using widgets to promote a money site. Gotcha Google.
Now, the one thing Cutts has mentioned before as being the major no-no of Widgets are selling links from them. Why is this do you think? My money is because it’s an obvious off topic link gathering scheme, where it’s hard to trace the origins of it, and the link is typically coming from a site that the person being promoted does not control. Sound good to you? Sounds good to me. Not for selling links, but for some more fun.
This entry is an untested one so far, but it was buzzing around in my brain so fast while I was walking home that I figured crap, I better write this down somewhere. Believe me, I would’ve tested it if I had a spare minute, but I’ve been really busy lately with some new projects. So follow my mental path here and have some fun.
Before Anything…Kill off the Footprints
Now remember that the entire loveliness of this tactic is that it’s hard to find the origins of it. So if we’re going to have this code distributed on a bunch of different sites throughout the net where people post the code to our widget, we want to make damn sure they can’t just footprint our script and find everything. So class names/div IDs and all that are going to need to change in each generated widget. Image file names? All unique. Domains where the images are hosted? Change those as well. It may be good to set up a script that automatically uploads an image to photobucket/imageshack every 10th time someone generates the widget on your site. That way you can’t get the widget found by G engineers looking for remote calls to your domain, and the bandwidth on the image won’t max out since you frequently have new URLs for unique ones. Any text in the widget should at least be slightly macroed/randomized so it’s not identical.
Don’t Make Waves: Distribute the Widget From a Variety of Origins
This is optional, but has huge benefits if it’s done properly. Sites that become popular, and have popular widgets are the ones that are going to get caught. So let’s say a normal, quality widgetbait site may get 5000 inbounds in a year from it’s widget(yeah, that’s a pretty popular one, but I’m sure been ones much more popular). A site that generated 5000 inbounds has a much higher chance of getting caught than 10 sites that generated 500 inbounds each.
So stick with me through this example (it’s a small scale example, but with a similar process it can get MUCH larger: don’t confine yourself to awards badges/widgets)
We create 3 blogs. footballblogawards.com, baseballblogawards.com, hockeyblogawards.com. Every 5 days, these things search Google blog search(automatically) and finds blogs that have to do with football or their respected sport(grab the 5 most recent entries, searches for a list of related terms in each of the 5 to ensure it’s consistently related to the sport), then it autogens an entry with a “blog award” widget that links to a site aside from our award site in the code.
If 20% actually put the widget on and 60% allow the trackback, and we list 10 blogs every week:
For one blog: 10*52=520 total awards given, 104 links to our designated external site, and a respectable 320 trackbacks scattered around the net.
Now let’s say we have 15 of these sites running in different niches. 15*104=1560 external links. 15*320=4800 trackbacks to our blogs. So we are now actually in possession of 15 relatively powerful blogs, and 1560 nearly effortless inbounds to our external sites. Even better, since it’s targeting blogs, chances are the “badge” or widget or whatnot is a sitewide link. So once again, let’s assume each site has maybe 150 pages indexed. 150*1560 domains = 234,000 total links. Mmmmm. Tasty.
Note: The numbers I use for this are guesses. The percentages I used are a bit optimistic, but using a fresh trackback IP every 4-5 weeks should make it doable.
But none of this matters if we don’t know how to link em. Hence the next section.
Deciding Where to Drop the Links
So we’re going to assume we’ve successfully removed most footprints, and that our sites are not going to get flagged by the big G. Our nemesis at this point are competing SEOs who would report our sites if they found out about our widget baiting indiscretions. Now, the big weakness of the eye check most SEOs do is that it only identifies/checks links that directly link to our “money” site. Why is this you think? They have access to our first 1000 links via Yahoo Site Explorer. Now lets say each of those domains has some 300 links on average. That’s 300,000 links for them to investigate. Not gonna happen. Beyond that, on a small scale(individual reports) it’s not worth reporting one individual, debatably spammy link that links to a site that links to us. That seems like it’s barely a boost.
But once we scale? It’s different. So let’s say we have just those 1560 unique domain links(more from individual pages) from the example above. Let’s also say we have maybe 50 different articles and social media links to our money site scattered around the web. Each of these 50 different articles not only have unique longtails probably not on our site, but they’re 50 unique opportunities to rank for the keyterms. So we’re going to rotate the spammy link in our widget amongst these 50. That gives us 31.2 sitewide links per article/social submission. After you multiply out how many pages that is, it’s a pretty significantly powerful ranker in and of itself(which drives traffic back to the “money” site), and in addition passes a lot of link juice back to us. Without directly implying our domain in any evil grayhat “link gathering” scheme.
If you stray outside of the “awards” area, there’s some definite hotbutton widgets that can be autogenerated and would be incredibly effective. Now, keep in mind this is not a bulletproof technique. A savvy SEO exploring the niche could find it and see through it. You could get footprinted. Any number of things could happen(though this is true for nearly any SEO tactic). But these are the ways I saw of staying under the radar while running these. That said, I’m excited to get this experiment under way(I have an idea not disclosed here that I think will be quite lucrative). Enjoy and experiment. Any results or concerns you guys see in this I’d be happy to hear in the comments.