Half-life of internet content

Some content has a half-life of one day.

For example, yesterday’s news article about water running down the middle of the aisle a Qantas plane; read once and it is unlikely to be ever revived anywhere for any purpose.

Other content has a half-life of a few thousand years.

An example is the bible; a set of letters massaged into a book over a period of 1500 years. The bible isn’t exactly a page-turner but it just keeps on giving. Eventually it will run out of puff or we will.

Indeed there are all sorts of content, with a half-life of between a minute (say a tweet) and a few thousand years, where the half-life is determined by how many people hear or read the content at some time period after when it was first constructed.

Noting that the concept of half-life is usually (but not always) that for exponential decay, I asked myself whether exponential decay universally fits the decline in use of all content. I suspect so. Initially there is a peak in viewing or reading driven by many factors, but which often looks exponential. After the peak there is a drop off in use, which can be very slow or quite rapid but in most cases might be described by a sum of exponentials.

However the bible is an interesting example – because the worlds’ population has risen, the number of readers may actually continue to rise over time (Google threw up no studies on this surprisingly!). But this increase in the world’s population can easily be accounted for by not using gross numbers of viewers or listeners but by instead using a percentage of the world’s population.

Therefore the half-life of content is defined as when the percentage of the population accessing the content has dropped to half of its historical peak value.

Does a soufflé rise twice? Not often … the same is true for content. By the time it has hit its half-life (as defined by the proportion of the population consuming it) my guess is that 999 times out of 1000 it’s on its way out.

If one were to measure the half-lives of various types of content then I imagine there are only 6 or so general categories of content:

1. Those that have a half-life less than an hour, e.g. a tweet
2. Those that have a half-life of around a day, e.g. the news story
3. Those that have a half-life of around a week, e.g. last week’s footy match replay
4. Those that have a half-life of around a month, e.g. a HBO episode downloaded via Bit Torrent
5. Those that have a half-life of around a year, e.g. a good e-book
6. ‘Pseudo’ evergreen content; examples are the bible, instruction manuals for life, on-line calculators, etc. I say ‘pseudo’ because all things eventually end.

I suspect that it would be pretty easy to pigeon-hole all content into one of these 6 categories at the point of construction.

There would be some exceptions of course, some content that surprisingly lives longer than expected and in the process attracts more consumers than expected.

The value of content is obviously related to how many people consume it and this can only be enhanced (or worse case unaffected by) a longer half-life.

It surprises me that people who invest their time and money into the creation of internet content construct anything other than high-quality evergreen content.

Unless of course the short-half life content is a loss-leader for longer half-life content or simply developed in order to finance the ego.