If you use similar tags to a high profile blogger, it can help with faster indexing and rankings, at least from what I have observed.

In many ways the effect is similar to linkfarms, using the tagging as a central hub. Many commentators complain when they see Technorati tag pages appearing in the Google SERPs, because Technorati doesn’t have a huge amount of editorial control over the content of those pages, they are more like search results.

In that situation however for any specific query, you are unlikely to have more than one or possibly 2 results appear.

Subdomains

In theory Google are meant to have changed the way subdomains appear in the SERPs, limiting how many subdomains can occur from a single TLD.

Matt Cutts addressed this in more detail

As mentioned on Search Engine Round Table, there would likely be exceptions to the rule for things like the large blog hosts, and Matt pointed out IBM as a likely search that should give more results.

So it would be perfectly natural for WordPress.com to be one of the exceptions, because multiple unique blogs might talk about the same topic, and give a good user experience, which is what Google strive for.

Tags As Subdomains

I notice on a query I used to receive occasional traffic from but for some stupid reason people are competing for it now – it rarely brought more than one query a day when I was ranking first, though it was targeted.

WordPress training in Google (note:- US geolocation non personalized)

The initial results are great.

One of the results is the tag page on WordPress for training – that is covered in my previous article which was questioning whether this is allowed – it is a tactic I would be keen to use if it was officially sanctioned, as would many blog networks.
Actually at least one blog network does do it.

Currently I use internal tagging and I know many of those get discounted as being poor quality, whilst others that are used more frequently are actually better results than any individual permalink page.

However further down the long-tail things get extremely messy.

When you study the pages, you discover that these subdomains were in some way intended for localized tags, so if your blog is in Polish, it would appear on a Polish tag page. Unfortunately the content on the pages is identical.

The English tag page for Training Resource could be looked on as a useful page, though it is still user generated content and easily abused, and a small webmaster doing the same might be looked on as a linkfarm.

The problem is that the Brazillian Portuguese subdomain contains identical content.

If that happens across for example 200+ different languages, then it represents a problem, especially seeing as WordPress.com is a commercial interest.

Whilst I think the original tag pages have merit, I am going to take a harder line with these subdomains.

If they had gone to the trouble of using a translation plugin to make the pages useful for discovery, they would have been useful, even though the translation wouldn’t be perfect. The title and headline being translated would make a significant difference on English search results, even if the URL remained the same.

But Automattic didn’t translate the pages, and they didn’t restrict these subdomains to only blogs written on that specific language.

As they might both read this, Matt & Matt, this in my personal opinion is search engine spam.

These are worse than your typical scraper site.

I am lowering the red flag

Carry on reading to find out why this isn’t the same as raising a white flag, and is far from surrendering to Google on paid reviews.

Robots.txt

I have spent a long time deciding on a course of action, and have decided that blocking my content using Robots.txt is ultimately better for me, and better for people hiring my services.

It also happens to be worse for Google than currently, but that is the beauty of this strategy.

It might be harder to rank, pages blocked using robots.txt still gather PageRank, and can appear in the index, though they would be looked on as dangling pages.

Ultimately links can always be redirected to a followup review which refers to the first, and that followup isn’t a paid review.

It is a little naughty, some people will sometimes receive editorial links within reviews and receive a trackback, but I don’t know of any spam plugin that checks robots.txt , plus the links will still be valuable in other search engines.

Google’s Achilles Heel With Paid Reviews

The only domain for which a client is paying for a review from is this one. When my content appears on other sites, there is a totally different editorial process, and links can in no way be looked on as paid links.

Content syndication is extensive:-

1. Social Bookmarking

Sites such as BloggingZoom encourage more than just a single line of description and rewritten titles on submissions, and not only deliver traffic from their existing user base, but also search traffic.

2. Hub Pages

Many content sites allow you to use syndicated content in the form of article feeds, and content is even picked up by larger sites such as Topix.

3. Authorized Syndication

You can arrange or organise for your content to be selectively syndicated on authority sites such as Andy Beard on WebProNews and even my WordPress SEO reviews published on SearchNewz.

Whilst I haven’t made it clear recently, I publish all my content under GPL, in fact I am switching to the GFDL with an invarient clause requiring a live hyperlink back to the original without nofollow – I prefer GFDL over creative commons because of this flexibility (for me) to be highly specific.

In future I am going to be actively encouraging syndication

4. Unauthorized Syndication

This is technically the same, but as long as people scraping my content are linking back to me, preferably with a followed link, it is great. I am not even worried about some light spinning of the content, as long as they state that the content has been modified and is only based on my original.

5. Indexed Search Results & Aggregators

This is the likes of Technorati, and feed readers that are indexed – I have no intention of blocking reviews from RSS feeds.

6. Multimedia

I use a lot of pictures and screenshots for my reviews, but this is going to increase – in addition I will also be creating podcasts and screencasts which will be widely distributed in their own right.

Hooray for Universal search!

No Nofollow = Editorial Backlinks

By not using nofollow in my reviews, it is most likely that syndicated copies of my reviews will provide backlinks not just for me, but also for my clients. The backlinks are editorial in many cases, someone has chosen to syndicate my content.

Unfortunately Google use backlinks to attribute content to an original source, but it is a whole lot harder if they can’t index the original. It will be interesting which site syndicating my work will rank highly, or how many.

Linking to Syndicated Content

This is something I haven’t decided on yet, but just like I can link through to my various social profiles, I do have the option to link through to my content on other domains after it has been syndicated.

Worse for Google

My content will still be in the index, filtered through an extra layer of editorial control, but there is going to be a whole lot more of it.

Google have made it clear that they are only worried about the existence of links, and not the time it takes to create content, expertise, and whether links within reviews were specified or given in an editorial capacity.

I honestly don’t like junk reviews written purely for SEO purposes, but as Google seem determined to impose the letter of the law rather than the spirit, throwing the baby out with the bath water, whilst I will comply to the letter of the law, I can’t see a reason why I shouldn’t sidestep the charging bull.

Nofollow is not the answer to Google’s troubles

Update

There seems to be some misunderstandings, and I need to clear them up.

1. The blocking hasn’t happened yet – it is the next thing on the todo list
2. I intend to get more search traffic from Google taking this action, not less.

Update 2

Robots.txt has now been modified

The list is quite short, but now I have a strategy in place, I will be writing a lot more paid reviews

Whilst this might be looked on as insignificant, some of those pages rank quite well for very useful terms, and are probably worth 2000+ visitors per month.

Update 3

Whilst the changes in robots.txt were quite straight forward, before making any reinclusion or reconsideration request, I thought it important to check the robots.txt within the Google webmaster console.

First of all I waited for it to be refreshed by Googlebot, which seems to happen approximately once every 24 hours.

There is an option to just copy and paste that refreshed data by hand, but waiting for it to be fetched is conclusive.

Next I entered in the URLs which need to be blocked by the robots.txt file, and checked them.

In theory Googlebot will now be blocked from crawling the “offending” pages, and I will be able to ask for reconsideration.

Photo credits
Lowering the Flag (modified)
Matador (modified)

Today while pulling an email from the spam folder I was greeted with this message.

This appears to be something newly implemented as Forrester Research were recently talking about problems with filtering.

There also seems to be lots of talk about an offline version of Gmail coming soon, but so far nothing substantial in the rumour mill.

Akismet – Proof of False Positives

Just today a reader using Akismet posted about me being a comment spammer. He was joking, but it was because Akismet had flagged one of my comments as spam.

• It was a brand new blog
• That was my second comment posted on the blog
• The first comment I posted was not flagged as spam

A Picture Tells A Thousand Words

7 Months Of Historical Data

• Most of my readers know I live in Poland, so that could factor into any Akismet calculations. I also use an internet service provider that uses a variable IP address.
• Most of the comments I have made on 100s of blogs over the last 7 months have been to this domain. A few have included my other public sites for wordpress plugins or google toolbar buttons.

With that amount of historical data I should be able to post a comment about any subject I like, and Akismet should give me the benefit of the doubt

Akismet

Whenever someone writes a list of plugins featuring Akismet at the top of the list, the blog spammers rejoice.

Blog spam is about eyeballs on their links, even if it is only in an admin panel.

Spammers even target things like referral stats for traffic, making you think someone has linked to your domain to go and check, and for old sites that still have top referrers.

Every time a blogger has to go trawling through their spam folders to check for legitimate comments, the spammers get eyeballs on their links, and with the primary content of spam being the most popular content and products on the internet, their market is not just the links from dead blogs, but also the eyeballs of active bloggers, many of whom are young males, who spend far too much time in dark rooms lit only by the glare from their computers.

Akismet will never succeed in killing comment spam until it totally removes the need of eyeballs on all those links.
It is a noble cause, and I am sure it will eventually be a successful one, but there seems to be something wrong with the collective intelligence, or certainly the “binary” spam/no spam black/white decision process, whereas Spam Karma offers lots of shades of grey.

How To Set Up Spam Karma

One of the biggest problems people face when they think about alternatives to Akismet for spam control is how complicated many of the alternatives are to configure.

I decided the best solution is to provide you with a crib sheet for Spam Karma, featuring the exact settings I use on my blogs.

These settings will result in some comments being flagged for moderation. Some spam will also get through, especially manual spam, and some trackback spam from splogs sending you legitimate trackbacks will also have to be moderated, at least initially.
Some of that could possibly be avoided with the plugins at the end of this article.

Those comments that do get flagged as potential spam end up at the top of the moderation queue highlighted in green, and as people comment more frequently on your blog, the chances of that happening to someone who is a regular is effectively zero.

If you want to add some extra layers of defense, Akismet can still be used as a plugin to Spam Karma, although this is an additional download.

Basic Spam Karma Configuration

Here is the exact basic configuration I use on my sites, some of which have been using Spam Karma for 2 years.

Download Spam Karma

Here is the basic configuration

Link to Full Size Image

With these settings I get nice emails every 8 hours for potential spam, those comments between 0 and -20 Karma – first time commenters with javascript switched off typically fall around -5
The first 2 or 3 trackbacks (with a link) from sploggers get through, and from then on they dig themselves into a grave never to appear again with the snowball effect, or I can totally blacklist them.

If you wish to also add some collective intelligence to the mix, then you can use the 3rd party developed Spam Karma Akismet Module

If you are using WP-Cache, it is advisable to also use this fix for Spam Karma with WP-Cache.

For Those That Insist On Using Akismet

Use the Antispam Collateral Condolences plugin from Mark Jaquith of B5 Media (announced a couple of days ago that he was joining). I can think of so many blogging network and business blog things that Mark’s talents could be put towards, and I might have to convince Aaron that he needs them ;)

Engtech’s Auntie Spam plugin for Greasemonkey which allows you to search through Akismet spam faster.

Alternative Plugins

Recently I have been hearing good things about using http:BL and the Simple Spam Filter
Some people swear by Bad Behaviour

I have seen lots of people test other combinations, but almost everyone I know who has an in-depth knowledge of plugins (independent of Automattic developers) and who has experience using Spam Karma ends up returning to it, possibly with the use of Akismet as an additional filter.

I was just reading Paul’s post on trackbacks as there seems to me some kind of meme floating around started by Steve Rubel that trackbacks are dead.

I smell FUD, because that probably ties in with the “Social Media Press Release” that is being promoted by Edelman as a solution for press releases (PR Web support trackbacks)

Trackbacks are very much alive for blogging platforms such as WordPress that actually pingback automatically without having to copy and paste some silly code that is supposedly to reduce the amount of trackback spam.

Trackbacks happen less on such blogging platforms – I have heard recently that Typepad only seems to be working to trackback other Typepad blogs – that is just sad.

Blogging without trackbacks is insular blogging

Anyway Paul mentioned receiving…

1000 Spams Per Day

Here are my stats for Spam Karma which I wrote about the other day

* Total Spam Caught: 9800 (average karma: -1056.58)
* Total Comments Approved: 1029 (average karma: 9.94)
* Total Comments Moderated: 419
* Current Version: 2.3 rc1

That is over 5 months

• I started this blog Oct 17th 2006
• I have been running DoFollow to remove nofollow from trackbacks and comments from the day I launched.
• I have been running Spam Karma since day one
• This blog is higher traffic (at least by Alexa)
• I have more toolbar pagerank
• I cover slightly different topics, but there is a lot of overlap

Paul’s domain is older by a fair amount but that doesn’t explain

15 times as much spam

What do you think?

Dr Dave has just released a new version

New in 2.3 of Spam Karma:

These are the full release details from inside the console, much more than is available on the blog. (for those that don’t have Spam Karma Installed currently)

New in 2.3:

* Miscellaneous bug-fixes (UI, WP21 compatibility, compatibility with other plugins etc.)
* A few small changes in the modules (improving blacklisting module, disabling RBL for now).
* Important updates to WordPress DB schema in order to try and improve its sluggish loading time (not directly an SK2 problem per se, but had to be done for SK2 to run fine).

Sorry this update took so long to come: basically SK2.2 worked well-enough for everybody and I am of the “If it ain’t broke, don’t fix it” school of development…
However, lately, a couple issues with WP prompted me to finally push a release out. More details on these changes for the technically-inclined will certainly follow on my blog.

A couple things in the meantime:

* While the amount of code changed isn’t very big, I highly recommend installing this update: it should dramatically improve both WP and (therefore) SK2′s CPU performances.
* Once you’ve copied all the files over the previous install of SK2, make sure you browse to the ‘Manage SK2′ page at least once so it can run the auto-update script.
* From now on, use this online form to contact me for any SK2-related issue (all previous support email addresses will be discontinued soon).
* Due to one really nasty bug in Apple Mail (no mistake on my end, I sweat), all the mails in my inbox for the past couple months have been deleted. If you contacted me in the past and I haven’t replied yet, please send again (using this comment form). Sorry about that.
* The Documentation Wiki has been offline for a while now: I had to disable it, as spam(how ironic, I know) simply wasn’t manageable and Mediawiki dearly lacks proper tools to deal with it. I am planning to run SK2′s documentation on a basic WP install instead, but will still need to do the import from the old wiki manually (volunteers anyone?).
* Despite insistant rumors (and low release frequency, I must admit), I am NOT discontinuing SK2 development in any way in the near future (full story here).
* Brian ‘ColdForged’ Dupuis has been working on a very cool SK2 plugin using mod_security. I need to check up with him, but the customized version I am running here is doing wonders, reducing server load by weeding out repeat offenders at the server level. BTW, any volunteer do work on a version using ip_deny instead of mod_security ? I swear it’s very little work…

Spam Karma is extremely low maintenance on niche blogs because if a legitimate comment does happen to trip the system, it is unlikely to have as many penalties associated with it as automated spam. You can set your email report function to notify you of potential good comments to a high frequency and a flexible value (I use 8 hour reporting and Skip spam with karma under -20)

I have some sites running on WordPress that receive a good 100 spam messages per day, and possibly 2 real comments a week, and I only log into them if the comments require a response.

Spam Karma is that good!

Spam Karma is among the currently 5 plugins nominated for my first round of donations to WordPress plugin developers from my income writing paid reviews. I will be opening voting on the first of next month, so if you want to nominate a plugin, this is your last chance.

TwitterAdder

• Automatically create Twitter sales channels
• Joins other sales channels in the same network
• Follows Other Users automatically
• Feeds affiliate links to sales channels
• Ability to take other people’s Twitter feeds and remix them
• Built in vocabulary from existing feeds “oh yum… pizza”
• Name Dropping
• Ability to randomly source stories from Techmeme plus associated blog posts

Software concept as per all my work is licensed under GPL

I have just read over on 10e20 a post by Tamar discussing how Twitter can be used for direct marketing, by creating a product feed of special offers.

Why is it OK on Twitter but not OK on a blog platform?

• Twitter is indexed by search engines
• Twitter content is syndicated as RSS
• Twitter is indexed by Technorati

The WOOT Twitter feed is effectively a splog by most definitions.

Why do it on Twitter? You could just subscribe to their RSS feed

There is a huge difference between this, and possibly commercial use of Twitter by their newly announced web services and command line, where you can request specific information, and have it sent directly back to you.

Lots of people are using Twitter commercially, pointing people in the direction of content on their blogs, or those of others. Twitter needs a way to handle disclosure.

Twitter is already ranking well in search engines, it would be so easy to create a linkfarm of Twitter Splogs.

MyBlogLog Spam

Lots of people are aware that there was a problem over the weekend with co-author requests being sent out, and some forced adding as authors to communities.

It was a long weekend in the US, but it seems the MyBlogLog team have already fixed it, rolled back any changes, and hopefully it won’t happen again.

I think this might have affected the larger communities more than the lesser ones. I only received a single community request and it wasn’t a forced inclusion

They have also posted a long description of their plans to control spam going forward. (via Darren).

I welcome the changes, and feel that placing “adding” restrictions will help significantly. We will see how the new messaging system works.

Widget Performance

I have seen quite a few complaints about MyBlogLog widget performance

This is normally on:-

• Blogs with a poor layout – If you load your sidebar first it affects your readership, and probably your search engine performance as well. Choose a theme where the content is displayed first, or shoot your designer
• High Traffic Blogs – I don’t think this actually has an effect on the speed, but owners of high traffic blogs tend to notice this more, and are more vocal about it

For those with a poor template design, whilst it might be a painful choice to make, the problem isn’t just the MyBlogLog widget – you really should have your content appearing first, and a good designer should be able to create almost any layout you desire placing the content first.

For the higher traffic blogs, there is probably a solution that could be adopted. In a recent comment on Jeremy Zawodny’s blog, Mark mentioned the following:-

Jeremy,
Just do what I do; a cron job that runs curl on the widget address every so often, saves the output, and just uses an include to display the content. I use a jsp include, but I’m sure you could do it in php as well. No relying upon third parties for your page display time performance, no annoying dns round trip times, everything just works.

Later in the comments I suggested that someone creating some kind of plugin that does this for WordPress users or maybe just a simple PHP include so anyone can use it is likely going to receive a fair amount of link love, and fix their performance issues with MyBlogLog and a host of other widgets.