1. Add the site to Google Webmaster tools

Obvious, huh? However, this allows tracking of Google+ button clicks easily and produces valuable data about click profile changes. This should be a basic step for each larger website anyways.

2. Fix the Facebook open graph tags.

Facebook’s open graph like buttons do not work properly “out of the box”. There are couple tags required for them to become useful and traceable. You will be able to contact page liker’s from within FB after the fixing the tags for home page. You should not forget to define facebook admins for the site in the web page headers. Then you will be able to add and monitor your webpage in Facebook’s business insights.
This tool will allow easy validation of each page: https://developers.facebook.com/tools/lint/

3. Add social events tracking to Google Analytics

This one is very important if you are using Google analytics. Although GA tracks Google+ clicks out of the box, it does not track other social events. So, you should add some custom JS in the page code to launch specific analytics events on social button actions. The generic code is to launch trackSocial method in Javascript:
_gaq.push(['_trackSocial', network, socialAction, opt_target, opt_pagePath]);
Only the network and action are required and they are passed as custom text. The opt_target parameter is useful when one uses like button for other page than default one. This page offers the best explanation how to implement facebook likes and tweet tracking : http://code.google.com/apis/analytics/docs/tracking/gaTrackingSocial.html
Note, that this code is for asynchronous tracker. It will require some modifications if you use synchronous tracking, and you should probably switch to asynchronous tracker anyway. Also, you should switch to new Google analytics interface to see the tracking results.

Fix the JS of sharing buttons to track social actions as well

The code above is not universal. It tracks clicks on the default sharing/like buttons only. However, you will not be notified after shares resulting from social sharing plug-ins like AddThis or similar. However, as long as you have access to API or can modify JS, you can track these events as well:

<script type="text/javascript">

var addthis_config = {
data_ga_property: ‘UA-xxxxx-xx’,
data_track_clickback: true,
pubid: “addthis-id”
};

</script>

This configuration is enough to start tracking social Addthis actions in google analytics. Similar approaches might be applicable by other social sharing services.

5. Consider installing Seevolution or other JavaScript heatmap tool

Some services cannot be tracked easily. In such cases there are one option only: track clicks directly. And the best tools for that are javascript heatmap trackers. There are some choices there, though Seevolution ( http://seevolution.com ) is great and allows tracking several sites for free.

6 (bonus). Add google Alerts for tracking your website name and brand names over the web

Not every social interaction originates on your site. You should monitor your mentions on google alerts and some other services as well. This might lead to perfect opportunity to get more exposure and new visitors.

Will these 6 things cover all social interactions about your website? Definitely not. However, implementing these steps will result in much better understanding of what is happening around your site.

Any additional ideas or comments?

Some of the “best posts” are ones that bash “Rogue” bloggers and sites with poor WOT reputation. I am on the bad side of this argument, as my rogue removal site 2-viruses.com has not too good reputation. This is how they attracted my attention in the first place.

However, the last post kind of crossed the line : Auto Blogs are Unethical and Downright Wrong . Here Frank claims that auto blogs are unethical, as they do not attribute the owner. And I completely agree that this is unethical, though simplifying everything to direct money gain is a bit too easy.

However, the problem with this statement is that he is not attributing the ownership himself. And that is easily proven by checking some things in the last removal descriptions provided on techjaws.com for Antivirus IS. Here a screenshot:

An interesting part on this screenshot is the way infected files are used. There are 2 lines referencing XP file locations, and one line referencing Windows 7 file locations in completely different format. Note, that there is an user User referenced, and the way random file names are used.

When I asked Frank on comments on the autoblog topic, I got this response, that he tested it on Virtual Machine. Well, that is possible, if he has VMs running 2 different operating systems AND uses inconsistent file naming out of hurry.

However, a simple google search clears all the doubts:

Note, in which company techjaws site is : All other sites are rated RED. At least one of them is referenced in other posts bashing sites with RED WOT rating. And all of other sites posted removal instructions for Antivirus IS couple days earlier than TechJAWs. are they mind readers?

To clear last doubts, here is screenshot from spywarevoid (I am not affiliated with that site, though I know people that work there and  that it does not belong to Frank).

Can it be that 4 different sites test Rogue on 2 machines in exactly same configuration and have same inconsistency displaying results, even if the sites are not owned by the same people?

Now I would not bash anyone for copying part of other people work. Even the attribution can be hard due to fact that you might miss the original poster, and do not want to credit someone that is clearly a rogue blogger. I am guilty of similar things in the past, though I am trying to do it better now. But I will never bash other people. And I will always attribute original author if reminded.

What I do not like is hypocrisy: bashing others and doing same thing. It is especially bad, as TechJaws displays advertisements (google ads and others), thus it tries to gain some money out of blogging.

Looking back, pageviews is a viable objective for some markets that sell main-stream products or try to strengthen brand. However, is it likely that I would go to a club in completely opposite side of the world or that I would do grocery shopping there anytime soon?

This issue affected display advertisement for a while already, and it affects social marketing even more so. People will not speak positively about unnecessary service that is pushed to them . More than likely they will laugh at it. And that will reduce credibility.

Back to real-world example. I have a marketer in my friend list that tries to push me a DJ fan page and events that are never close to my location. The marketer found me by inviting fans of other DJ’s that play internationally and reside in my location. Do your research, people. I see no reason to become fan of page that is not relevant.

How to remedy situations like this? Invest some time into looking for related local groups. Local clubs the DJ plays would be good start instead of international DJs themselves. That would lead to fans that are relevant and (maybe) interested. Simple, eh?

Earnings are not necessarily the most informative metric of the website. Some websites do not earn money directly. The same amount of money earned might suggest good or bad website performance depending on investments done, visitor count, market or advertisement type used.

It is even more difficult for websites whose income sources are advertisements or which advertise main business indirectly (for example, using branding). The conversion might be long process and you see increase in sales or conversions, but it is hard to estimate the performance of content.

However, you have to analyze the content. There are information that can be extracted from both statistics and user-generated content related to the original content. These things might not correlate!

For example, we post articles in our local fashion magazine and publish excerpts in facebook pages. Most viewed content not necessary is most commented one! Why it is so ?

1. Content that appeals to main auditorium of website is more engaged. Content that is related to auditorium is most interesting.
2. Shocking content would be more viewed one. They are interesting for broader spectrum of people.
3. Shocking content has more viral potential as it might be redistributed by non-members. That is why single-time visitors are useful as well.

Is engagement more important that visitors? For membership sites, definitely. You grow your crowd and your influence in them. However, a single non-related re-share of content might increase your site visibility further than in-door engagement. Thus in many cases you have to vary your content enough to attract more diverse crowd to your site.

Spending time in some forums and social boards I met couple types of people that give professional advice there: people that were in same situation, contractors and affiliates or products. The most of people would argue that the last group is most annoying and they can not discern between contractor and other two groups. However, I would like to slightly disagree.

First, a security contractor is a person that is on payroll by specific security software or service company and gets static amount of money for his job and/or bonus on how well company is doing. An affiliate is a person that works for oneself and gets money from specific amount of product sales. Quite often affiliate has more than single products he offers. So, what is the practical difference? Here are some myths:

Myth no 1. Contractor’s quality of advice is often better as many of them have better knowledge in the field.

Partly true, most of contractors work in the field. However, it is not true for hired marketers compared to security experts that refuse to work on contract bases and earn additional income from sale.

Myth no 2. Contractors have stable income, so they do not need to force each sale

Not true. If contractor does lousy job, he will lose his income (contract) completely. Doing good job might yield a bonus. If an affiliate does lousy job, his profits will diminish. However, lot of affiliates try selling as much as possible because they seek profit.

Myth no 3. A contractor does not need to use sneaky tactics at promoting product

Completely not true. Contractor has additional benefit at using sneaky tactics because they can pretend being former customers and there is hardly anything that would prove it otherwise. They do not need to use tracking codes, they do not need to disclose anything. They spamming techniques might reach borderline.

Myth no 4. An affiliate will not promote best product because he is out for profit

That is again not true as you can’t promote BAD product for long. Additionally, affiliate has a huge benefit of being able to choose what product to promote and what not. A truly good affiliate is not forced to promote a product using false comparison tables, spam or by accusing competitions business model.

Sure, there are all kinds of affiliates and contractors. However I would check such things as:

1. Accusing other reputable products being bad because option x is paid one
2. Leaving user without a choice when choosing product
3. Calling other marketers spam because they promote different product than they do
4. Calling others a scam because they do same things and are paid by performance
5. Avoiding giving free information and pushing a product
6. Forcing user to pay for free things to boost sales of product without giving them to try it
7. Making a set of disclosure rules others (affiliates) have to follow to be legitimate but not following them themselves.

I got a lot of examples of such behavior. And sadly it will remain that way in the market.

And there is one thing I have to disclose – I was a contractor of security company for a short while too (No, it was not Malwarebytes, but the experience would be the same I guess). I did not liked the experience, because my freedom to promote right product in each situation was reduced. Now I can give suggestions of any product I think is good for the customer. Even if I get 0 cents from it. That is freedom I have working for myself and not being a contractor.

And if you think it would be if I would work in different company, that promotes “best” “free” product? Well, think again. I will have less freedom and less choices.

Now I do not say that being contractor is evil. Being contractor and not disclosing it when suggesting a product is evil, though. Much more evil than being affiliate, which is easily seen in most cases. Calling other legitimate products malware is evil when you are paid by competitors. Calling others spammers when you do the same is evil.

• No obvious optimization
• Content is crappy, mentions term once and in very different context
• Few or no links
• Design is old and ugly

However, such site is in the first page and sometimes quite high in results. So, what are the reasons for this?

Personally, I have seen a blog that ranked really well under one porn term with single mention in blog post. It had nothing to do with porn industry though.

Sure, we do not know Google algorithm, but my guess is disambiguation of search results plays major role in niche markets. Quite often the key phrase is not enough to determine searcher’s intent and search engine has to provide alternate meanings for the term. Sure, in many cases the Nr1 is site that performs best in terms of SEO, but there is room for different sites as well.

Let’s discuss an example. A person searches for “online poker” – a quite competitive term. In the first page we will see poker rooms, poker information portal, poker forum, some news sites and images. It is not because forum sites and poker rooms are at the same level in terms of SEO. We get these results because Google cannot determine which result is most relevant for us:

• We might want to play online poker in top rooms (highest possibility for the term).
• We might want to read reviews of these rooms
• We might want to discuss about online poker
• We might want to read some related news, etc.
• We might want some local results as well.

Sure, we will not see poor sites under this query in top places. Online Poker is too competitive. However, some of the sites might appear out of nowhere even under this term.

So, what can we conclude from this?

1. Copying top sites is not the fastest way to front page. Unique approach is.
2. The longer the query the more SEO work has an impact.
3. Behavioral analysis and local data will have an impact on what results you get as well. They matter. Guess what sites target users need, and not what they use now.

Thats all today

1. Your server and domain
   1. First thing I would check is the spammy IP list. Is your IP address blacklisted? It happens, especially on shared hosting. Good place to start searching is http://www.mxtoolbox.com/blacklists.aspx . I would use a dedicated IP if I plan to do mass mailing.
   2. Do your IP address resolves back and how? If it has IP address in reverse domain address, they think it is a DSL – assigned IP and not a hosting one, so it is assumed that this is SPAM. If IP resolves back to your domain, it is a good sign.
   3. SPF record is highly advisable. Not all incoming mail servers use SPF record; however it never hurts to have one.
2. Your email message headers. Forging headers is bad idea. Your headers should be informative and correct, including message ID. It is important how you write your email address there as well, you should always use real name with email address where possible. Same is applicable for recipient email address as well. Prove that you know them.
3. Avoid spam words or keep their density low. The message should have enough text; however it should not be too big. Also, avoid excessive capitalization and spacing.
4. You should include information why they are getting this message, also your contact information, as well unsubscribe information (if applicable). You do not have to write that your message is not spam – only spammers do that.
5. You should use well-formed html if you have to use it. No hidden text, tracking images or similar – they are usually blocked by anti-spam programs. Also you should include text version of email.
6. Avoid using spam – mailers. The anti-spam programs know about them and how to recognize most of them. This is quite sure way to be flagged as spam.
7. Monitor your returning email box. See what messages got undelivered and bounced back. See if it is your problem or the recipient problem.
8. Upon registration, ask your visitors to add your email to their address book and use same email for communication with them. Address book is the best way to pass spam filters.
9. Some mail servers do not like too many emails from single sender at any point. Try to space out the sending of your newsletters.
10. Avoid sending too many messages to people that can’t read them – full message boxes, blocked, etc.
11. Check your web application if it can’t be used to send spam for real. This might be a case for “send to a friend” forms or similar which allow information input. This is quite serious security issue as you might get blacklisted.

Overall, the most important factor is monitoring back all information you get back from emails you send, especially the “negative” one : bounced emails, sending errors and similar. There are lots of information in bounced messages that you should read and evaluate.

First, this involves Google reader, and if you do not want to switch to it, it is not for you. Google reader has a “share post” option which allows to share content to your followers. Even if you have no followers, you can use this feature, because you get Atom feed.

This feed can be integrated into your Hootsuite account (settings ->RSS/Atom ->Add feed). Set time frequency to 2 hours, and limit post amount to one – you do not want to spam your readers with content like many automated twitter accounts do. It is different from automated approach, as you review the content you submit to hootsuite, thus you take care not to submit older news or uninteresting posts that happen once in a while at any feed.

Sure, this will not replace tweeting manually. You do not get all content through RSS, do you? You should take your time reading other people tweets and sharing valuable ones further. Also, not all sites provide RSS feeds or are worth subscribing.

Hootsuite (Hootsuite.com) is a very good web-based twitter account manager. It is very feature rich:

1. Multi-user multi-account management. Very useful for corporate twitter accounts.
2. Multiple columns, tabs for columns, custom groups of users. This is quite important for bigger accounts. In fact its interface is quite similar to tweetdecks or Seesmics, thus I got used to it quite fast.
3. Tweet scheduling functionality, which might be important for some
4. Feed integration (similar to twitterfeed)
5. Ping.fm integration for upgrading fb and other accounts.
6. Own url shortening service which is quite reliable and provides statistics. Also adsense integration in it.

However, there are some issues in Hootsuite that I would like to be changed or fixed:

1. Separation of groups from tabs/columns. I might want to put same group of people in multiple tabs (as I need to see them more often).
2. Better FB integration. I use Seesmic for reading FB as well as I don’t have time to check website often. Full integration with stream API would be very very nice.
3. Some things bug in groups. For example, tweets of some people are excluded when I add them to groups even if they are in main view. This is because some users do not appear in twitter search. However, this information would be easily recoverable from home stream. Or I do not get a reason for separating groups from keyword search.
4. Custom color schemes. Yeah, I am picky.
5. Ability to clean up column once it is read, or a way to distinguish between tweets that are unread.

Will I use it? Yes, definitely. Hootsuite is a very nice software and I see lots of potential in it. However, I cannot abandon desktop client in hootsuites favor yet.

All regular twitter users have seen the face of Bill Crosby and couple other super twitter get rich schemes. Their followers promote ways to get thousands followers quickly and then get rich by turn-key systems. You don’t even have to watch after your account, everything you have to do is sit and wait for revenue.

I do not believe these systems and that’s why:

1. These schemes rely on selling twitter traffic products to other users and nothing else. The more stuff you sell, the more diluted market is and in the end only owners of the program will prosper.

2. Furthermore, tools for automatic follower increasing decreases value of each your follower because there is a larger percentage of other automatic twitter account. Same bots like yours. Lot of people will not read your tweets thus you will not earn anything.

3. You are not special in selling this product – everyone is selling them for same price. You have no distinguishable product to sell. Would you buy a product that gives you almost same followers the person that sold the product got?

4. There is no good way to feed interesting content automatically. You can feed a news source, but there is no good reason you would be more interesting than original source. If you feed more content, you will spam your followers. Sadly, you need to generate content by hand.  You need to create content yourself.

5. You have no relationship with your followers.

The single way to be on social media is to participate. Without participation, it is just advertisement on other websites. Sadly, that’s how many marketers understand “social marketing”.

There is couple of reasons to develop site distributed under multiple subdomains, e.g. subdomain1.domain.com, subdoman2.domain.com, etc. Each subdomain might be a distinguishable mini-site, which offers better separation of topics and even protects main site rankings to some extent.

However, subdomain approach has one issue that needs to be solved. That is each subdomain is treated as separate site by some search engine spiders. This leads to insane amounts of concurrent hits. The worst offender is yandex – a Russian search engine whose spider scan list is probably alphabet based. Thus many of your minisites get scanned at the same time for content.

This might not be a problem if each of your minisites are hosted separately and not on the same server. Also, this would not be a problem if your subdomains are simplistic. In other cases you have to rely on your page caching and server capacity.

There is a crawl delay option that yandex listens to. However, this option can be set in robots.txt only. And robots.txt is subdomain-based. Thus it works for single subdomain only.

I have decided to totally block yandex from indexing one of my sites once. Luckily, this site can live from visitors from Russia, as it is geared towards different location. But now I doubt I would do a heavily multi-domain (300+ subdomains) site knowing that I will have to use yandex traffic as well.

I am affiliate of one unnamed program that is a bit confusing for me.  The software in question is sold worldwide, however you get affiliate sales for purchases made in USA only. This troubles me a lot, as the product is good and I use it myself.

The question is what difference it makes where electronic goods are sold to? I see couple of possible answers, none of them too good:

1. Trying to compete in single market only or trying to conquer specific market. But each additional sale is good and useful to the company. And yes, you are getting them despite I will not get money from them.
2. Internationalized support.  It is good to be concerned about the users, but how this is related to the sales? Yes, it is common to have different commissions for visitors from different countries. I am ok with that too. Or are visitors from UK less valuable than ones from USA?
3. Saving money by paying for part of the traffic. Enough said.

Oh, and if you wonder if I know that someone had paid… My friend made a purchase apparently, as he believes in the product as well. Interesting enough, it is not notified that this program for USA only in their account at CJ.

Should I stop recommending that product as well? Well, I am strong believer at recommending only products that I believe in, no matter how bad their marketing is. So, for now I will not stop recommendations.

However I think that in long term they will lose market and trust of people that promote their tool, as the web is global. One should care about one, centralized affiliate center for all their affiliates no matter where sales will come from.

I use Google Ad Manager for showing additional content in one of my sites. It is quite good tool, however there is a significant drawback for it: it has no real handling of CPA and shared revenue based offers.  CPM and CPC is everything you get. And sometimes you need CPC ads to compete with CPA offers.

As an example, you have single ad space in the page. You might want to distribute CPM offers through the month evenly, but also show CPA and CPC ads at the same time. Also, depending on current eCPC, you have couple of competing affiliate deals.

Sure, you can solve this problem by entering estimated CPC value for the ad. However, this workaround is not too good: CPC you get from affiliate ads is visitors-dependant often and you might not get same CPC as other sites. Thus initial CPC value is only estimation of real value you will get. You need to update the CPC of these ads continuously. And if you test out various ads for various regions, it is quite a time consuming task.

I wish that Ad manager would release some kind of API that would allow automation of such task. That is posting and changing eCPC values for various ads. This would allow ad revenue maximization for the site by combining all (or almost all) approaches. Is it hard? I do not think so. It would require more dynamical ad rescheduling, but in the end all parties would win. Google Adsense would compete with CPA ads and it would allow Doubleclick integration.

At the mean time, I am planning to take a better look at OpenX project. I think tt does not allow CPA ads either, however I could tweak its PHP code accordingly. I am really interested to see how effective its code is as well. There are some additional benefits of OpenX as well, for example Google-independence, though I am not too concerned with this for a moment.

1. Google webmaster tools. Adding your site to google webmaster tools is good idea, and there you will get information about links not found. It is most simple way, but worst as well. This is because Google detects the links that are linked and indexed. However, there might be plenty other links that need to be taken care of. For example, advertisement campaigns use non-followed (often javascript) links.

2. Log analysis. That is the best method if you have access to log files and you can process them. I prefer using Awstats for that, however you can do it by hand for smaller sites on apache, as errors are logged to separate file as well. The single problem with manual analysis is that error log has less information than common log. There is no referrer link mentioned in the error log. However, this can be solved by using grep to scan access log for 404 error codes as well. The drawback is that some CMS processes all requests and do not generate error codes successfully. This means that you will not see such errors in logs even if they exist.

3. If you can’t access error logs, the best way of action is to use custom error pages and create a log from them. You have to log both referrer path and request uri for best result. This approach can be implemented in many of the abovementioned CMS’es too.

So, what to do with bad links? This depends on what causes these links. If it is an advertisement campaign or a referrer site, you will have to create a redirect from bad link to the appropriate good one. In cases this is a malformed comment link, I would just delete it from database.

I use Google analytics for site visitor monitoring like majority of webmasters today. It is free, reliable and quite good in many of the respects. However, there are couple particular cases where you can screw yourself and analyze its results incorrectly. Here we will discuss one particular problem that is not too common, but happens quite often: by default, Google analytics handles multi-domain website visitors incorrectly.

Google analytics is geared towards having separate accounts on each subdomain. This means following: A visitor coming from one subdomain to another one is counted as completely new visit. This inflates your traffic ranks (effectively doubles them or more). This might happen even in cases if there are no redirects between versions without www and with www. Additionally, pages with same uri on different domains are shown as one and the same page in analysis. This is not as bad, but undesired behavior, as it makes harder to track everything on your analytics accounts.

Note – there is no real traffic decrease, however the traffic is shown 2x higher before fixing problems with subdomains.

Sure, inflating your search traffic might seem handy when selling sites or bragging about daily visitor count. However, it will not help you in the long term.

How to solve it? I have used partial solution for quite a while, where you add custom filter to put subdomain information in uri. However, it did not solve the problem with visitor amount. Recently, I have found a full solution.

You have to add a custom filter first :

• Name: Subdomain_fullurl_hack
• Type: Custom filter – Advanced
• Field A -> Extract A: Hostname > (.*)
• Field B -> Extract B: Request URI > (.*)
• Output To -> Constructor: Request URI > /$A1$B1

Secondly, you need to modify tracker code to set domain to your domain.  This is done by adding a line after your tracker is initialized (below var pageTracker = …. line) :

pageTracker._setDomainName(“.yourdomain.com”);

Btw, thank you folks at searchdigital for this guide!

If you want to check the results, I suggest comparing ones you get from google analytics with ones you get with log analyzers, like awstats.