This blog is NOFOLLOW Free!

Archive for July, 2010

Twitter as a marketing tool? It sucks!

I really don’t understand the recent twitter fever. Twitter is one of the most useless and stupid internet websites ever created. Put a message in 140 characters and throw it away. Ok, so? WTF is this?! Why don’t we use a SMS for this?!

Ok, ok, the idea is that, unlike the SMS’s, tweets can be seen by anyone. I understand this. And having said this, I understand also that Twitter is a tool for voyeurs and exhibitionists, isn’t it?

I have 4 Twitter accounts, all business related. The best of them has ~300 followers. These 300 followers are useless, and let me tell you why. We can express the interest that a follower manifests in your tweets by a ratio between his/her accounts and the number of the accounts that he/she follows.

Let’s suppose that you have 2 followers, John and Jimmy. They both have only one twitter account. But John follows 30 accounts and Jimmy follows 5000 accounts.

We can express their interest towards your tweets as follows: John has an interest of 1/30=3.33%, while Jimmy has an interest of 1/5000=0.02%. Your tweets are basically non interesting for Jimmy. In fact, Jimmy is not interested in any of the accounts that he’s following. On the other hand, John has a little interest in your tweets.

Now, make an exercise: go to your Twitter account and browse your followers. See how many people your followers follow. A lot, isn’t it? Simply put, your pity tweets will be lost in an ocean of indiferrence. As you get more followers, your message will become more and more lost in this ocean.

The only legit use of such stupid tool it’s in publishing vouchers and coupons. Yeah, this works pretty well, but only with big brands that can have something massive to offer. Dell. Vodafone. Nike. Apple. Etc. Global brands. If you have a little translations company and you need clients, don’t waste your time with a Twitter account. It’s useless.

W3C validation SEO talibans ask you to validate your code

The newbies think that Google is a beauty contest, not a search engine. And this is obviously false. The truth is that Google doesn’t pay a cent on tags and validation, but it pays attention to:

  • the content
  • the ratio between content and code (cleaner code is better code)

So, the validation doesn’t help with SEO, despite all the buzz that validation talibans are trying to create recently.

Don’t validate your pages, it’s useless. The HTML language it’s an alive language, yesterday we were using HTML 4.01, nowadays we are using more and more HTML 5 elements, including elements that aren’t yet in the official validation specifications. For example, a few years ago people used images for rounded codes. Now we can use the border-radius property, with browser specific extensions: -moz-border-radius and -webkit-border-radius, that will never validate.

Instead of validation, focus on:

  • Cross-browsing code: is your website readable with every big browser (Internet Explorer, Firefox, Chrome/Safari, Opera)? No, do not believe what the validation talibans are saying: a valid code will not display¬†always in the same way on every browser. Just check with the “margin” and “padding” properties and see how they are displayed on Opera and Firefox: it’s different, no matter the validation.
  • Clean and semantically clear code. A good code will make the website usable both on computers and mobile devices. Search engines will understand better your content. Clearly define your content areas and don’t mess the <divs>.

In fact, all this validation buzz is a pathetic marketing strategy of a bunch of losers. They have nothing of real value to offer, so they must differentiate somehow from:

  • the talented designers, who are offering to their clients good, functional and clean designs;
  • the good SEOs, who are offering to their clients good backlinks and valuable advices;
  • the functional coders, who are offering to their clients clean templates and a good code/content ratio.

Stop this “validation” buzz. Web languages aren’t the Holy Bible. Don’t be a taliban, it’s pathetic.

Google Adsense Interest Based Ads: what’s the problem?

I started recently to investigate a little bit the “Adsense interest based ads” feature. The reasons behind my investigations were two:

  • First of all, I have observed in the last months a drop in my CTR and consequently a drop in my revenues – but I didn’t take it seriously, believing that this i’s only an effect of the crisis.
  • Navigating pretty much on a website that we work on, I observed that the “interest based” ads were very unrelated to the pages’ topic; moreover, the so-called “interest” was wrong: for example I was browsing, few minutes before the visit, on some web hosting related webpages, without any desire to buy something, just to research, and than Google system decided that I would be interested by web hosting ads, which was wrong!

So, I decided to investigate further this problem and I discovered that a lot of people that have targeted, content rich websites, have the some problem: the interest based ads just overpass the contextual ads and this leads to a lower CTR. Interest based advertising seems to work well only on unfocused websites (for example social networks or general discussion forums).

Even on this blog, I observed that the Sciphone related pages were displaying only 10-15% Sciphone or chinese phones related ads and 85-90% unrelated ads, which was totally unacceptable. If I’m browsing now this page, it means that I’m interested now on this topic!

On the other hand it’s true that interest based advertising can display – as Google says in their Adsense Help pages – the highest paying ads. However, this is useless as long as you don’t get clicks. Better fewer bucks and more clicks, than many bucks but no clicks.

So what are the problems that arise with the interest based advertising?

  • The interest itself is contextual: if I’m browsing now some “web hosting” pages and 10 minutes later some “plumbers in calgary” pages, then my interest 10 minutes laters relates to “plumbers in calgary”, not “web hosting”!
  • Google is unable to guess the real interest of the user, and the real scope of an user’s interest: was it to buy? …was it to gather information?… was it to help his old mother to gather some online information?… was it to help his old mother to buy something?… was it the user itself, or a friend using the computer?… etc. This “interest based advertising” it’s just a lottery.

How could Google improve its Adsense technology?

  • First of all Google should let the advertisers decide at a website level where to show interest-based ads and were to show only contextual ads. Now you can opt out from interest based advertising at account level. This is no good, because one can have focused and unfocused websites under the same account. The publisher should be able to choose to activate or de-activate this option for every website.
  • The “interest” based ads should be activated or de-actived even at channel level. One website can have highly focused webpages or unfocused webpages. For example, the homepage of an “yellow pages” website is very unfocused, however the inner pages are very focused. On the homepage the advertiser could open an “interest” based channel, while on inner pages he could open only contextual channels.
  • Google must improve the “interest” calculation algorithm. For example, if one has a single browsing session related to a topic, this is not a clear sign of “interest” for that topic. But if it has 5 browsing sessions, yes, we definitely can say that the user is really interested in that topic. If one browses one single day on a specific topic, this is not a sign of clear interest. But if he browses 3 day out of 10, yes, this is a good indicator.

My 2 cents on this.