What to expect in SEO in the coming months [video]

I would have though this is a no brainer for Google to catch these days. I though text under pictures would be like white text on white and would set off immediate alarm bells.

Google has never been good at handling short term black hat spam (buy viagra is the colloquial example always). They are good at filtering sites out after they've been ranked for a while though.

But when these blackhatters are making tens of thousands of dollars a day with their short term rankings, it's hard to convince them that they need to be focusing on the long term rankings avenue...

Comparing a "big" site like that to a small business who might have $10-20 spent against them to negative SEO doesn't really compare.

It's like comparing McDonalds to a local burger joint.

I remember a comment by Eric Schmidt that the best way to combat spam was through "brands". While brands are a component of good search results, I can't think of a more misguided philosophy. If the only advice Google has about where to buy something is to buy from "Amazon" then Google has lost its utility in revenue generating searches. I think the best way to combat SEO spam is through engineering and hard work. Giving all the power to established brands is a shortcut that will ultimately undermine Google's core value to its users.

The incentive doesn't disappear, as there is a great deal to be gained by being bumped up search results. They raise the costs or circumventing the system, but that just means people look elsewhere. Its an arms race.

http://xkcd.com/810/

Now the function of a link has changed entirely. Many of them don't even exist to be clicked, and many of those that do are either advertising or the functional equivalent.

There are a few places where a link really is still an endorsement: for example if I put together a hacker news comment that links a blog post I think is informative. Or a (genuine) blogger that writes about a topic.

However, anytime there's a useful signal to be found in something like that the SEO jerks come and flood it. That's why we've got comment spam, we have reddit sock-puppets, we've got blogvertisments and so on.

And the so-called white hat SEOs aren't much better in this regard. They are pushing their clients to go write mediocre "content" that is largely duplicative of what's already out there. People should only be writing content if they have something -- preferably interesting and/or informative -- to say, not because it means more people will visit their brochureware site, and certainly not as a means of getting (tricking?) people to click on no-value-add affiliate links or AdSense ads. That's putting the cart before the horse.

By this do you mean popular people will rank more easily like they do now? How is popularity determined? By links?

re: How do you solve it?

You don't. At least, not how Google does it. Google is trying to act like the government of the internet and tries to manipulate behavior into things that align with Google's goals - primarily making money via AdWords. To achieve this goal, they try to discredit competing advertising paid link systems because they break Google's system. It's akin to twitter making it all but impossible to have a successful twitter client.

At some point, there is too much money in the current system to render what Google's webspam team does as much more than an elaborate game of whack-a-mole. They hit spam, more spam pops up elsewhere.

The only way for Google to make their algorithm better is to take out or dial back the "buyable" elements - the social ones like links, tweets, facebook fans, etc. They did this pretty effectively with AdWords over the years. It's harder with organic results, but not impossible. Arguably, for many people you could put 10 reasonable, relevant, non-spam results on the front page of a given search term and it would be "good enough" for many people. Something akin to DuckDuckGo.

Since this is a non-trivial problem what Google is doing is basically better than anyone else so far, and there is a lot of value in the social signals if only because it makes people care more about Google than they care about ranking on Twitter or Facebook. So, kudos to Google on what they're doing, but there is nothing inherent to search that requires a link to be a vote or for that to be such an important relevance factor.

Most of the collateral damage is going to be people straddling the grey hat line and rightly so in my own opinion.

My site is pretty much all user generated content (car reviews), and Panda seems to struggle with differentiating such sites from content farms, unless they're part of a larger, established brand.

Almost two years of attempting to fix the problem by improving quality and layouts, reducing ads etc had no effect. Then I raised my case on Google's own forums, resulting in a lot of attention, and a few weeks later I saw a massive improvement (I know, correlation is not causation). Unfortunately I seem to have been hit again last week, by something that looks suspiciously like Panda (though I'm not 100% sure).

I've mostly moved on (now working on an iOS app), but it did appear from the outside that Google was comfortable with the impact of Panda, and it's good to know that they're still focussed on improving it.

If your in a competitive market where you have to worry about negative SEO so much so that it is making an impact on your SERPs then I'm sure the web spam team would be more than interested in hearing from you

First point on googles webmaster guidelines says it all. Your chasing google, build a site for your users first, google will follow because it's in there interest to provide the most relevant content.

  Avoid tricks intended to improve search engine rankings.
  A good rule of thumb is whether you'd feel comfortable explaining what you've done to a website that competes with you, or to a Google employee.
  Another useful test is to ask, "Does this help my users? Would I do this if search engines didn't exist?

This is clearly nonsense when some webmasters were buying thousands of links.