
Risks and opportunities
197
relationship management practitioners. We offer an adaptation for use in
online public relations.
1
So what does uncertainty look like in the fast-moving online environ-
ment? De Meyer et al offer four uncertainty types: variation, foreseen un-
certainty, unforeseen uncertainty and chaos.
Internet traffic data, displayed in time series (a sequence of events, meas-
ured typically at successive times, spaced at o�en uniform time intervals
such as daily or monthly), have a number of characteristic properties,
widely known as ‘stylized facts’, which are different from other kinds of
time series:
They tend to be long-tailed; in other words, there is a higher frequency
of very extreme events that have a long life and tail off slowly.
They tend to show long-range dependence; for example, search engines
will find content that is old and present it today, people will remember
and bring content to the fore long a�er it was news to another group and
so forth (the internet has a ‘long memory’ and ‘time-shi�s’ information
– and reputation).
They exhibit volatility; in other words, the apparent variance (from the
plan or anticipated outcome) is not a constant but tends to fluctuate
irregularly, something the internet has in common with traditional
media that can bring back old news to support a new story.
These are challenging concepts, but can be visualized by thinking about
books or recordings sold through an outlet such as Amazon. Traditionally,
a band might release a single that was bought contemporaneously by a lot
of people, making it a ‘hit’, before it slipped out fashion. Traditional record
shops stocked the big sellers and knew it was not cost effective to maintain
stocks when the song dropped out the charts. But the economies of scale
offered by Amazon, allied with the infinite amount of virtual shelf space
it commands, mean that songs that most have forgo�en are still available.
OK, they will sell in very small quantities but, they are sales nonetheless
and the cost of storage and display is minimal on the website. This is the
well-noted ‘long-tail’ effect.
2
The same long-tail effect can be applied to news stories. Whereas once
most people would read a story when it was splashed across the front
page of a print newspaper and then discard it, such stories are now part
of the digital archive, quickly accessible to search engines. An individual
story may not be read by a huge volume of people on any one day, but its
readership stretches down the long tail (the ‘value’ of the story has a very
long ‘shelf life’). This too has implications for PR, in that today the news
never quite goes away. It may be forgo�en by most readers, but Google has
a long memory and is always ready to serve up scraps that organizations
might imagine had long vanished.