January 31st, 2008 sam
I enjoy developing software, little tools to solve problems that I have. For example - I created a little tool to read Bid Buddy (now the rebranded Tradedoubler Searchware 4) bulk URL outputs and work through each URL and search for 404 responses in the server returned http header. If anyone is particularly interested in this, i’ll happily e-mail it out.
A screenshot of Social Spam Buddy’s interface
One of my more interesting projects was a tool I developed at my former search agency. Essentially this tool was designed to spam social bookmarking sites with client bookmarks under several hundred accounts. I believe there are commercial solutions out there, but an in-house solution is always the best option if the time scale for development is small (the tool took about 15 hours of total production, minus documentation).
10 bookmarks * 10 bookmarking sites * 100 accounts = 10,000 back links.
(at a post rate of about 2-3 per second single threaded)
This turned out to be highly successful for Yahoo and MSN, but had a negligable effect upon google rankings. I’m not entirely certain on the results though, I just coded it and sent it to the optimization monkey’s. The back link figures are particularly impressive really, however, social bookmarking sites are an obvious target for spammers due to their very nature - they are essentially just huge link farms. Additionally, the api’s that several of the sites have (delicious, furl, spurl etc) make them ideal targets for software such as mine.
I made several interesting additions to this software in my own time, that I chose to never push live or allow into the public domain.
The first was the introduction of TOR ip masking. Although slightly fiddly - you had to install the TOR software, and then my software would interact with it via telnet, it worked - albeit significantly slower. I never pushed this build live for two reasons;
- Ethically, I find the use of the onion router for commercial gain and spamming to be extremely distasteful. I think that project is extremely important for the more repressed regions of the world.
- I always hated spam, it would have made it far too successful and realistically difficult to track.
The second was revolving and auto changing text, to hide the visibility of the spam. For example; anchor text would stay the same, the url would remain the same but the text of the description would be modified slightly in order to hide from spam detection. Combined with IP masking, this was simply far too powerful in terms of evading detection. I doubted the spam would ever be found realistically when combined with the final addition to the software.
The final unpublished option was the simple spam management option, this would post the bookmarks at random times to random accounts throughout the day (or any arbitrary timeframe that you desired). Coupled with the previous options this just made the spam far too successful.
Essentially, prior to my involvement the company had a large team of people spamming these links manually throughout the day. Now - realistically, is this project blackhat? Most would say yes, although during my work on the software I didn’t stop to consider the implications of my work. Had it been more appreciated, would I have given out the completed solution? Possibly. I do believe that action such as this signals the death of the social bookmarking websites that don’t adapt to fight spam. They are basically dinosaurs now, and this is reflected in the google ranking of the links themselves - regardless of sites such as delicious having an excellent pagerank.
Yahoo does, however, remain important regardless of the rejection by Google. As delicious (incidentally, one of the easiest bookmarking sites to spam) is more and more integrated into the Yahoo search platform, will the backlinks that are created in this online solution become more valid within the engine? Only time will tell - currently they purport that they have relatively no value, but I find this hard to believe in an industry where successful assimilation of multiple platforms (read: google and its many wings) is the situation that is aspired to.
Posted in general ramblings | No Comments »
January 31st, 2008 sam
So the big news in SEO recently is the huge hit GoCompare have received due to their link co-op use. This behaviour is typical when these insurance affiliate sites are worth so much money, so having your site rank top for ‘insurance’ or ‘car insurance’ is worth millions.
An excellent example of extremely recent link spam can be found with moneyexpert.com, whom have almost 10 million in bound links all from the interesting digitalpoint co-op! So what can google do about this technology, which can easily provide millions of links and essentially linkbomb a site to top position?
- Penalize sites that advertise on the network, rather than the sites that run the adverts.
- Modify the google robot to include a link positioning rank, so that if the link is in the footer of the page it has less value than elsewhere on the page. Studies into a website based on something like crazyegg could help.
- Stiffer penalties for companies that engage in blackhat seo.
Ironically, the digitalpoint co-op is clearly related to the ‘greyhat’ social link directory spammer that I developed although considerably more powerful..
Posted in general ramblings | 2 Comments »
January 30th, 2008 sam
I’m a search manager for an media agency, I manage accounts and deal with clients to do with the search spectrum - something that I hold more than a passing interest in. My first exposure to search was running a small adwords campaign several years ago, prior to my undergraduate degree. In a professional capacity, I started in an entry level grass roots analyst role and worked from that role to an executive and beyond. This time has given me experience across almost every online marketing vertical, most of which were extremely competitive in both click cost and conversion achievement costs.
I’m young - only 24, but I’ve always been involved with the internet since an early age and i have always wanted to be on the cusp of an emerging technology in order to be able to ride the exciting changes and problems that this early uptake could introduce. It is clear for anyone to see that online marketing and in particular search marketing is still extremely immature and direction for advancement is clear by taking components from various sources. The great thing about search is its current state of constant flux, that someone who operates outside of it for six months is already a proverbial dinosaur in the field. If you are entering it now, outside of the basics - be wary of what people teach you as fact and “best practice”, particularly if they haven’t had search exposure in the past 12 months. “Best practice” is one of those brilliant misnomers that is consistently miss-used. “Best Practice” 12 months ago meant spending £50,000 a month on some accounts that can now barely shift a tenth of that at a acceptable rate. Things change, engines change and the search world moves extremely fast.
What I find most interesting about search is the way in which the market essentially calculates its own value based on the profit margins of the companies involved. The jostling for position and attention by advertisers certainly mimics something that you might expect to see on an Attenborough documentary. If you could significantly lower your bottom line costs for your product, you could potentially engulf your selected portion of the search market - although this high level of change and spend attitude is not something that has happened in my experience yet.
One of the beauties of Pay Per Click is the way in which it discards popular marketing paradigms in its transparency when it is employed correctly. You generally know exactly how well that it is working, and if a desirable level has been achieved - pushing it for superior performance is possible. Currently, pay per click is treated in a different way to that of regular advertising and marketing media in that its accountability is considerably higher and it is kept on a significantly tighter leash - however I think this attitude is likely to leak across into other mediums as the assimilation and conversion of technologies continues to occur.
The purpose of this blog is really an information and technique sharing device, detailing some of the more important rules of pay per click and changes within the field. I will elaborate and clarify upon the themes mentioned above..
Posted in general ramblings | No Comments »
