The Fresh Egg blog

Latest digital marketing news

What Does Bing's warning against SEO Link/Like Farming Really Say?

Gez Hebburn

Duane Forrester published a post to the Bing Webmaster blog on August 31st titled Link farms and Like farms – don’t be tempted showing how easily Bing could detect and disregard artificially manipulated social media signals.

Specifically, the post looks at the distribution patterns of genuine social activity compared to that of 'link farms' or 'like farms'. Forrester asserts that artificial social media activity is far less organic, chaotic and natural than the real thing. In Duane's words, "in most cases, if we spot like/link farm activity, we simply ignore the signal."

In most cases? Interesting get out clause there... reserved for paying customers?

If this warning is more than just a bluff, we should be wondering why the link & like farming services currently being offered in some of the internet's less reputable alleyways are working at all. Most of the SEO world is convinced that a simple and honest social media content-propagation strategy encourages faster indexing, drives traffic and influences SERPS - effects which appear to scale proportionally with the magnitude of the social activity.

Is Bing only detecting the most blatant link/like farming, and presenting a simplistic version of the data as scare tactics? Wouldn't anyone attempting to game social signals professionally be able to create a more sophisticated pattern of distribution, one that modelled an organic response far more accurately than the artificial pattern Bing presented?

Diagram apparently showing artificial link farm growth
Image from Bing: artificial social growth

As explained by Forrester in his post, the search engine's formulae for identifying undesirable farming tactics were bound to be more complex than just clocking up 'likes over time'. But what other metrics are being used?

Identifying individual users and their networks will come into play, but how much of that analysis relies on IP addresses to identify unique individuals or potential link/like farms? If they are relying heavily on IP data then Proxies, variable IP systems and carefully scheduled, exponential spamming are possibly all that a farm operator would need to create responses that look a lot more like Bing's organic graph.

It's worth remembering that plenty of creative entrepreneurs have been 'artificially manipulating' social bookmarking metrics for a very long time, so this isn't a new phenomenon we can attribute to the current social media darlings, Twitter and Facebook.

In this context at least, likes are just a variation on good old social bookmarking - an indication of an appreciative audience keen to share content with their networks. The 'gamers' utilise extended networks of real people and bots, some running multiple accounts, sharing and promoting their web content through a mixture of public, secretive, natural and paid networks. Despite the best efforts of search engines to marginalise and negate the effects of what they see as undesirable activity, bookmarking manipulation and associated networks are still effective - and presumably mutually beneficial for those involved - after almost a decade.

Image showing natural social growth
Image from Bing: natural social growth

Just look at some of the garbage that gets into Google news and alerts, and Bing's equivalents. Look at the mountains of no-added-value price comparison engines and affiliate sites that are still profitable. The visibility and relative business potential of so much low quality web content stands as testament to the fact that dodgy tactics which impersonate a caring audience still work, disproving Forrester's assertion that Bing's got this all figured out.

If the search engines really were clever enough to negate the effect of artificial social metrics, farming wouldn't be a growth industry.

Come on Duane, tell us more - only half the cat's out of the bag. As a result it's easy to suspect that where identifying social signal manipulation is concerned, Bing - and presumably Google? - may be small dogs with a loud bark.



Share this post


comments powered by Disqus