Thursday, July 29, 2004

Theory of Blogilution

This is a long and convoluted post. It is a technical post. It may require some degree of abstract thinking. Essentially, it’s me trying to understand something for myself and sharing conclusions with the rest of you. If you don’t really want to hear a theory about life, the universe and frequency of posting – wait until my next post. Suffice to say the below outlined theory simply serves to justify my reasons for deciding not to post every day anymore.

In the last week, some things have flourished:

...and some things have not.

*looks around*

Regular readers will know that I decided to up the frequency of my postings from once every four-ish days to one every day.

Unfortunately, I think that the quality of my posts has suffered, so I'm going to go back to the old frequency and quality – and try to slowly bring up the frequency from there as I become a more efficient writer.

This decision was made when I realised that q>t for my blog and that I should have one metric as dominant to the other – always.

Confused? Well, I DID warn you I was a business geek and a processes and systems geek. I will now also reveal my great love for microeconomics graphs in bringing you The Blogosphere According to Monica.

You see, I've had a good look around and have observed that blogs fall into some very broad categories. (I'll mention the ones I read often, as I think I can comment on those more readily.)

There are the massive aggregators such as Instapundit and Vodkapundit, the 'this is interesting/this is strange' commentators such as Survival Arts and World Wide Rant, the theme commentators such as Samizdata, the 'welcome to my life, kitchen to your left, bathroom is down the hall' journals such as Belle de Jour and the 'pondering the mysteries' navel gazers like me. I realise there are others, but these will suffice to illustrate my point.

For all of these, frequency is a factor in audience generation and retention. It is the gauge of effort and of health of the blog, it is what entertains and what keeps the material relevant. But you see, I believe that as you go down my list, frequency becomes less important and the quality of the writing becomes more.

I've formulated a personal set of frequency/quality theories, where 'Quality' (q) refers to the complexity and richness of writing demanded by the readers (note: this isn’t really a value judgement where high ‘q’ is better than low ‘q’. Fitness for purpose is important, some blogs don’t NEED to have lyrical prose choking the important information) and ‘Timeliness’ (t) refers to the importance readers ascribe to getting the information NOW.

Both are measured from the reader’s point of view, where I tried to gauge what percentage of a reader’s satisfaction would be sated by each respective factor. I presumed* that where timeliness (t) was a low percentage satiety factor, posting less frequently would be permissible – where if it were a high percentage satiety factor then posting as frequently as possible should be a goal. So far, so obvious.

*(I think my old uni lecturers can probably sense this going on and are gnashing their teeth nicely. I haven’t actually received a research grant for this idea. I haven’t even a PhD – and here I am making PRESUMPTIONS! Where will it end?)

Essentially a similar presumption was made for quality. Where quality (q) was a high percentage satiety factor, the readers may be willing to give the blogger more time to formulate the posts. Where quality was of little concern, frequency of postings should be higher as the reader would gain value by the variety and volume of posts and the links that were provided – there would be an expectation of more postings more often.

These kinds of linear relationships lend themselves to graphs. So I graphed them. Both graphs shared a common axis (frequency of posting) so I put them together. I may be arrested by the Academia Police for creating relationship diagrams rather than just memorising them for exams. If you don’t hear from me for weeks, you’ll know what happened. The diagram below is a visual representation of the q and t lines I was talking about. Click here for a larger version so that you can actually see what I'm going on about. 

You can see by the dashed lines that I took the time to plot four blogs to illustrate my point. They are:

I = Instapundit
S = Samizdata
B = Belle de Jour
T = Th’ inkwell

Let me go through the four types of examples to explain what on earth I’m trying to get at with all those pretty colours.

Two caveats:

1 - The quality of the links, the choice of topics to write about and the general tone of the blog is outside the scope of my little measurements. Don’t try to subsume any of these factors into the ‘q’ measurement.

2 – The observations noted about these blogs are generalisations. I realise that not all posts to these blogs will follow my rules. Blame it on that annoying ‘free will’ stuff you hear about.

The first blog I thought of plotting was Instapundit. I had noticed that the posts were very frequent there and that the blogger himself usually added little original commentary to the link and its quote (where a quote was supplied). If I were to guess what his readers log on for, I would say that they come to ‘graze’ through his offerings for something interesting. I would suppose they expect a high frequency of posts but aren’t overly concerned with the way he expresses himself. In other words, very high t and low q.

Then I thought of Samizdata, which also has many posts a day but the posts have an article-like structure about them. This isn’t just about pointing the reader to an interesting link on world events but posting some pithy commentary about the topic and shaping an opinion around the link, extrapolating it’s pertinence to higher level abstractions and general political trends. So what do readers expect of Samizdata? I would say that timeliness is still an issue – especially where links and comments are made on the day’s political manoeuvrings - but leniency has to be given to allow the blogger(s) to scribe something of value around the event. The quality of writing is more of an issue than with Instapundit as readers are interested in the wording of the commentary as well as the link provided. Here, I would say that t>q as for Instapundit, but the gap is significantly narrower between the two.

The next blog I decided to place on the diagram was Belle de Jour, a classic journal blog. Journal blogs help us to step into the life of the writer, so frequency of posting is important as we come to expect to be in their confidence regularly. Just like reading a chapter of a book at a time, few of these posts give us a complete story by themselves, but are at their best when read together to give us what we want – an overall picture of the person or of their profession/life/relationships. Because the posts need to be read sequentially, we get impatient if there's a break – hence the need for frequency. The quality of these posts is extremely important, there's no point in reading the journal if the life in it is expressed poorly or tediously. Here, I would say that q>t, but the margin between the two isn’t exceptionally wide.

Lastly, I thought about Th’ inkwell and where it would fall in the whole scheme of things. The purpose of my writing is to explore an idea, research it where relevant, draw a conclusion and present it as best I can in a flowing, informative and entertaining form. Here, the ideas themselves are timeless – current events can be used as a starting point, but are often no more than that as the writing delves into the deeper and more abstract issues behind the triggering event. Timeliness is important to let the readers know I've not been hit by a bus but I think that people log on for the quality of the writing above all. I would argue that I have a very high q quotient and a low t.

And here’s where we get to the group activity part of the diagram. By plotting where you think your q and your t values are on the x-axis, you can scoot over to the y-axis and obtain two readings for suggested frequency. Although the scale on the y (frequency) axis was chosen entirely arbitrarily, I still think that it does give some idea of where in the frequency continuum you might want to settle.

The diagram, therefore, points to me settling for lower frequency. This is a good thing, a very, very good thing. Firstly, it means that I can become reacquainted with M. Secondly, it means I won’t cringe when I look over old posts that were cobbled together as a last-minute thing. Thirdly, it shows that I am entirely capable of creating theories that prove what I want them to prove. It might be time to start postgraduate work.

Something interesting I noticed when I drew a horizontal line across the intersection of the two graphs (as you do just for fun) is that I could split blogs into what I call 'external focus' blogs (where the blog looks at, comments on, links to and revolves around things going on outside the universe of the blog and its writer) and 'internal focus' blogs –(where the topic of blog posts is the writer and their thoughts or experiences.)

As I mentally plotted other favourite blogs onto my diagram, I realised that the trend of distribution above and below the line by external/internal focus continued (for my exhaustive sample size of 0.000000000000003% of the web).

All blogs above that line are ‘externals’ and of the t>q nature, meaning that frequency of posting is predicted to be high to satisfy the audience. I wonder if this is because these blogs are highly reliant on their outside links (It’s no fun to hear about something interesting and follow the link to find the information is no longer there. It’s also not particularly interesting to hear old news with little commentary.) and must constantly refresh posts to stay relevant.

Right, those that are still cognizant after that little onslaught are no doubt wondering what happens when q=t. Well, honestly, so am I. Perhaps, if q>t=Internal Focus and t>q=External Focus, then q=t=Crosseyed. I do know that it would be hard to keep the readers happy and difficult to make a value judgement over which to sate where a choice between either t or q has to be made without one being obviously a more dominant driver of customer (reader) satisfaction.

The second purpose of this thought exercise was to decide on the metrics I would use for my blog. I knew that both t and q were important, but which was more? The conclusion that q>t for Th’ inkwell has helped to clarify which is primary (quality) and therefore which should be ‘anchored’ to the right level before I start changing things to improve the other (t).

It also boils down to the fact that the posts I most like are the ones that took a couple of days to think about, a couple of hours to write, were allowed to ‘breathe’ for a day and then edited for another hour or so. I can’t produce them at the rate of one a day without having a life support system hooked up to my office chair and a gravity well distorting space-time around me so that:

a)      I can have flashbacks to old episodes of my life
b)      I can, well, you know – do the high frequency, high quality thing.

Of course, I could have just spent two days mapping out the topography of my navel.


Please only use comment system below


Weblog Commenting and Trackback by