Webtropic

The norms of twitter behavior are pretty much user generated. @ Replies, retweets, hashtags to name a few, were all invented by twitter users as a kind of behavior first, and then were turned to application functionality.

Such a kind of social behavior is Follow Friday too: each Friday twitter users tweet to their followers certain people as worth following. Such of tweets are tagged (or hashtagged, to be more precise) with a #followfriday tag (or its shorthand: #ff).

By suggesting people one feels are noteworthy and have something interesting to say and share, one gives the other twitterers the chance to discover new people to follow. If they do actually follow is depended on the trust and influence that a suggester can inspire and exercise.

Yet, I always felt reserved to get into the habit of #ff.

Why?

Simple: I follow people either because they are real life friends and acquaintances or because they have something interesting (to me) to say.

For friends and acquaintances, I  have no need of suggestions and  I do not see how could I recognize whether those suggest by #ff as interesting to me.

Over time, the only reliable method that I have come up with in discovering new people to follow, is by observing other twitteres’ retweets: if in the retweets that show up in my stream, I see someone appearing often and I do not follow her/him already, then I have a prospect.

Observing the frequency of appearance of a certain twitterer in retweets  over time, I make sure that I appreciate hers/his style, conduct and  interesting content. This way, I know before I start following, who and how is the one I am to follow.

The method is not infallible, but is far better than #ff.

One might argue, that #ff is not a science or a detective clue, but a form of social reward for those we really appreciate.

This is very true, and it is a generous and welcome gesture in this respect.

But then the same applies for retweets. And while a single retweet does not say much about the person retweeted, a lot of retweets say, well, a lot!

In conlusion: do you want to recommend me for #ff? Retweet this!

Say a twitter account is listed x times.

So what, you might ask.

Indeed, the absolute number means little.

But by observing  both numbers as a ratio (followers to lists, F/L for brevity from now on), I think I have found a quite meaningful use of  them.

If  a twitter user is performing some sort of broadcasting of news, offers, weather, traffic, stock prices etc, it might useful to follow. But doing so in the timeline breaks the convesation and de-humanizes the stream.

Following such users/accounts through lists is much more convenient, much more practical. It brings sanity to the timeline.

If my assumption is correct, then such users/accounts would be listed relatively more often, than followed. Hence, the F/L ratio  would be relatively lower.

This metric, if true, can be used for identifying and classifying an unknown (to us) user.

My guess is that if the ratio is less than 10, then this user/account  is probably a broadcasting account of some sort. Or his activity is such (: publishing own blogposts, without engaging in conversations).

I have two personal twitter accounts: one for english and one for greek. The F/L for the first is 16.75.  Surprisingly, the ratio for the other account is approx. the same 16.62!

What is your F/L ratio?

I have been a big opponent of the vanity twitter use (aka harvesting followers, hoping that “followers” equals “audience”).

In practice, this meant that from a point on,  I completely stopped looking who is following me, did not reciprocate at the cost of being perceived as arrogant and kept my follower/following ratio to 4.

Still not content, I unfollowed quite a number of twitterers (some of them pretty big names)  on the grounds that they were either producing too much noise, or were talking about things I found irrelevant to me.

For over two years now, I keep experimenting with twitter:

• In the beginning, it was conversations. But as people kept flocking around twitter, conversing became hard, if not impossible.
• Then it was news tracking which, although useful, it was far from complete. Yes, the news came to me, but not the news I was always interested. And with it came a lot of repetition and nonsense.
• Then, based on retweets, it was content discovery and evaluation.
• Occassionaly, it was polls, mini-crowdsourcing, asking questions etc
• Grouping people allowed to create filters: filters for news, for content, for community info.
• Finally, there came mindcasting. The most interesting use of twitter. The one I subscribe.

The grouping feature offered by many twitter clients, has, for a long time, being the single organizing factor that brought some order into chaos.

But, lately, we have another one, far too important: Lists!

Although lists look pretty much as the  groups of twitter clients, they are not the same: groups are for  the people we follow or those that follow us, while lists are for everyone! This difference is a game changer.

Already people use how often they are listed as a measure of importance, influence or popularity.

But lists have another function: they are metadata. The criteria we use to classify twitterers in lists, describe what they are or how we view them.

Also, lists, unlike groups, can be public, can be viewed and subscribed by others. And as such, they bring focus and attention from another angle.

“Ok”, you might say. “Lists bring new features. So what?”

Lists can bring back the sanity in twitter. They can undermine the follower fallacy, they can bring value to ordinary users as well as to businesses and marketers.

How?

By allowing us to make a fundamental distinction: following is an action of trust and, to some extend, intimacy. Subscribing to a list is  willingness to be informed.

So if you are on twitter to spread your message (be it news or offers or corporate messages) seek to be listed, not followed. Your very intention implies that you most likely want to use twitter for broadcasting and not for creating relationships. That is fine. You won’t have to pretend you are a ‘friend’ from now on. You aren’t. You never were. But  now message spreading can be done without undermining the everyday experience of ordinary users.

—– end of part I —–

My lists…

Image from Flickr by RBerteig

I remember a time (not so long ago) when people were disgusted by partial feeds and most of the major blogs and bloggers abided by this unwritten law. Publishing a partial feed was synonymous to cheap exploitation, putting ad revenues over readership, alienating your readers etc.

I, myself, am guilty of accusing blogs for  this kind of ‘malpractice’.

But things change.

I see more and more probloggers publishing partial feed and very few readers  complaining.

Why?

Because the way we read and follow blogs has changed too. Information is coming to us. And usually it comes in the form of a short sentence with a link in it, something especially true for twitter.

So the trend of reading everything in the coziness of our feed reader is in decline, while reading from the source is back with a vengeance.  Twitter and Facebook are the benefactors of blog ad revenues.

A couple of days ago I came across this terrifying presentation from John Graham-Cumming.

Although the topics covered weren’t entirely new to me, put together in one presentation, had an impact.  I came to wonder if and how would the major web 2.0 sites work, if javascript was out of the picture.

I decided to make a little test to find out: I disabled javascript from my browser  and started logging  in such sites to see how would they behave.

Here is the outcome for the three most important for me.

a. Twitter

Most of the functionality was in place: the timeline, friend and followers. From the various buttons on the tweets and the timeline pages, the reply did work but not the fav button.

The direct message and delete buttons did not work either. Same with the drop down where you select a follower to dm, and finally, the followers and trending topics buttons.
But all these are rather trivial. Because most of the tweet buttons replicate user behavior (putting the @ sign in front of another user name for a reply, or the d letter for a direct message).
Not being able to fav, or, more importantly, to delete is a loss, but not a major one.

b. Facebook
Things are worse in Facebook: while Home, Profile, Friends and Settings are accessible, the inbox and chat are not.
Also, from the bottom bar, the applications menu is inaccessible. Most of the edit links and buttons don’t work either and finally the status updates, link sharing , photos etc cannot be submitted.

c. Youtube
Here things are disastrous: without javascript you cannot see the videos! On top, you cannot access your account settings or you mailbox. There was no point looking for more.

A small gallery with pics of the failure areas of the above web applications follows

The recent update  of twitter’s  Terms of Service, brought to my attention this page from twitter support :  The Twitter Rules. Is it not a long read but it is quite educative as to what twitter considers as spam or spamming behavior.

It is interesting to note that there is no rigid definition of spam:

What constitutes “spamming” will evolve as we respond to new tricks and tactics by spammers

Instead, the following  14 points list of spamming behaviors is cited.

• If you have followed a large amount of users in a short amount of time;
• If you have followed and unfollowed people in a short time period, particularly by automated means (aggressive follower churn);
• If you repeatedly follow and unfollow people, whether to build followers or to garner more attention for your profile;
  
• If you have a small number of followers compared to the amount of people you are following;
• If your updates consist mainly of links, and not personal updates;
• If a large number of people are blocking you;
• The number of spam complaints that have been filed against you;
• If you post duplicate content over multiple accounts or multiple duplicate updates on one account
• If you post multiple unrelated updates to a topic using #
• If you post multiple unrelated updates to a trending or popular topic
• If you send large numbers of duplicate @replies
• If you send large numbers of unsolicited @replies in an attempt to spam a service or link
  
• If you repost other user’s content without attribution.
• If you have attempted to “sell” followers, particularly through tactics considered aggressive following or follower churn.

The one  in bold has a special interest.

If your updates contain mainly links  you are considered a spammer!

Well this is news!

There’re thousands of accounts in twitter that do just this. How? By linking a blog feed to a twitter account. In such a case all twitter updates are links back to the blogposts.  Leaving aside the fact that these might not be appealing accounts to follow, considering link-posting as a spamming behavior  contradicts the presence of major media organizations in twitter and there are no signs that twitter actually objects their presence.

But if link-posting is ok for, say, CNN why would that  be bad for blog xyz with the 20 followers? The rule becomes a size discrimination.

Another notable notion in this rule is that twitter seems to still attribute value to the personal updates.  For me personal updates are irrelevant but that is not the issue. The issue is that one should have the right to write about the things he truly cares. If drinking coffee with his spouse is one of them, that’s fine. But if not, that should be fine too.

Besides, in order to have personal updates you must have a person too. With all these business accounts, what sort of personal updates is to be expected?

Maybe the reason why Twitter succeeds is because people don’t really want to have conversations. They just want to be able to scream out into the void and listen for echoes.

says Victor Ganata.

If true, then all web 2.0 product developers should go back to the design desk.

The real underlying question though is: “What is a conversation?”

Web 2.0 is about conversations. The markets are conversations, says the Cluetrain Manifesto. New services want to be conversational. New marketing urges us to form relationships and interact through conversations.

With so many claims  on the term ‘conversation’, I am afraid the term is stretched to a point where it will either break or become meaningless.  Defining  is confining.

If we accept that ‘conversation’ is a more serious kind of discussion (as opposed to chat), we can hardly apply this notion to  the conversations happening online. Most of them are simply chat.

Next comes the shouting in the void, like the tweets  ’I woke up and I am drinking coffee’ which occassionaly turn into a chat again.

In blogs, we often see  large threads of hundreds of comments, which, excluding spam and trolls, can be deemed as real conversations but not as one conversation. They are mostly conversations between the blogger and the commenters and secondly between the commenters themselves.

The pingback mechanism has  enabled a more sophisticated kind of conversations:  through blog posts. These can be extensive and spread to too many blogs, so they are difficult to follow.

In Friendfeed,  humongous threads are commonplace, especially if Robert Scoble is the initiator.  Yet, I don’t know many that  read such threads from start to end. So, what is the point of these threads, conversation wise (because I can think of many other points apart from conversation)?

In real life, you can have a conversation with only a few people. You cannot have a conversation with a whole football stadium!  Likewise, online conversations that can have an impact, and feel like they do,  are the ones  that people can participate from start to end, understand who else is participating, and catch up really quickly, if  joining late.

For this reason (and contrary to the popular perception)  twitter and twitter-like tools, by restricting the length of the what is being said and by limiting the participants of the conversation,  turn out to be more conversational (subject to abuse though).

When  people complain about twitter not being conversational, they may actually complain not been allowed  to blah blah endlessly.   But we agreed that this isn’t a conversation, didn’t we?

These little javascript programs can mean a lot to personal productivity and social media engagement. I came to this conclusion only recently, while I am using bookmarklets for years now.

The oldest I can remember of is the delicious bookmarklet. Honestly, without it my delicious account would be empty.  There is also a firefox add-on which, combined with the buttons does a lot more, but, hey, it is simplicity that we are after, aren’t we?

With the advent of Google Reader Notes, ‘Note in Reader’ became   kind of a regular but never really got off. I use it to share things in Reader occasionally, but most of the time I share things either in twitter or in friendfeed. Why? They show up much faster and the crowds in these two services are so much bigger than Google Reader. Google Reader is something like a library: you can’t make too much noise, so its perfect for reading systematically. But when you want some action this isn’t the place.

‘Press This‘ has been around for quite some time now. Maybe it is the second bookmarklet that I have come accross. Unless you want to blog about news, or respond to posts you have seen, it is not that useful. And adding pictures and media seemed hard. For some reason I had’t used it much. Until two days ago when I discovered, that much like posterous and friendfeed bookmarklets, it allows you to easily pick from the pictures of  page you want to blog about, and put them into you post. For thematic blogging it is superb. And deprives you of none of the wordpress blogging  amenities.

For twitter I do not use bookmalets to post, while there are quite a few out there. But I use one to shorten urls with urlborg. Then I manually copy and past the short urls in twitter. Why? Because I am using too many different clients and can’t have a unified solution.

The Friendfeed bookmarklet is a must, provided you are on Friendfeed, and you better be there. Makes it so easy to share a piece of content and spice it with an image or two, that it is irresistible.

Posterous was a late discovery of me. While  aimed at posting through email, does have a bookmarklet too, with editing capabilities  of the content shared, plus an extra area to add your own thoughts- comments. The combination speeds up casual blogging and, actually tempts you to blog more.

One thing I would like to see, is

One mark to rule them all, One mark to find them,

One mark to bring them all and in each browser bind them

to paraphrase J. R. R. Tolkien.

Yes, there are more. Much more. But the above are  the ones I care most. How about you?

Conventional twitter wisdom has it that being retweeted is a measure of value. This sounds very reasonable as

• being mentioned (=retweeted)  is a form of recognition and entails visibility as it, usually, includes the twitter name of the originator
• retweeting distributes a piece of information virally
• and, therefore, retweeting  is used as a measure of one’s twitter status by many twitter ranking services

But, let’s take a closer look :

There are two elements in every tweet (those retweeted included, of course): the who element and the what element.

The who element is the source or the channel  of the information compressed in the 140 characters.

The what element is either the tweet itself, or the content it points to. Usually it is the latter.

Retweeting tends to preserve the what but not the who. And even the what most likely will  propagate  changing form (=pointer) in the process.

After two or three retweets  (X retweets Y, who retweeted Z etc)  the very size of a twitter update makes it necessary  to drop some of the @ references: If Z is not in the social graph of X, but Y is, then X will most likely keep Y in the retweeting chain and not Z.

Also, if the retweeted message contains a short url, there is a high probability that some retweeting down the road, one will be tempted to use his favorite short url service and not the original. Let alone those that will drop the retweeting completely and present the information as their own discovery!

There are two cases of undeniable benefiting:

• If the tweet retweeted is someone’s statement and not a pointer, the retweeting process will most likely preserve the reference to the originator.
• In all pointer cases, the content pointed benefits from the incoming traffic, regardless which  short url service was used, if any. And this is great, because it makes twitter work like a social bookmarking service. Retweet equals Digg!

But where is the personal value derived from retweeting? And how much is it?

Being a reliable information source within your social graph is acknowledged by retweeting, but the opposite is not necessarily true: retweeting some random piece of info that one comes across does not yield value to the originator. At least significant one.

What I am trying to say here is this: worthy content will surface in Twitter and benefit from it. People will not benefit equally. So chose wisely where you put your investment in this service.

Screenshot by imjustcreative

A basic one at least.

With the advent of the second gen twitter clients, which support, among other things, groups, users are confronted with higher barriers to entry and exit: In all clients, the painstakingly prepared groups are hardwired in the client. No easy way to get them out. When one desires  to switch to a new client, he has to recreate all these groups by handpicking users one by one.

The problem becomes apparent even in the case where one does not necessarily want to change twitter client, but he simply has to work on two different (or more) computers.

In my case, I had to recreate the Tweetdeck groups for my desktop and laptop computers. And I did it only for the two out of four computers I use and for two out of seven operating systems (3 windows, 1 mac, and 3 flavors of linux). I did not even manage to create them as exact copies.

Now, this call might sound like a luxury request, but given the path the twitter clients have taken (check Nambu or AlertThingy or Seesmic Desktop, to mention a few), you will notice that ‘groups’  is one of their prevalent characterists. So dealing with this feature effectively is essential for the success of the product.

And it is very simple really. A CSV or XML file would suffice. With bare minimum information. The name of the group, the twitter account it belongs to and the follower or friend ids that belong to it.

And even if the software vendors do not to go this way for fear of loosing users, there is something else they could do, to allow portability of groups for the same client but different computers: store this piece of info in the cloud. Create a simpe web app where each client will connect to and retrieve the group data.

Some people have  already hacked their way to group migration for Tweetdeck but this, clearly, is  not the way for the masses. Hence this call. But is anybody listening?