It was news
to me (until Professor Kagan told us) that bots can act like humans and so are
even harder to detect than if they just acted like machines. I knew that fraud
(and ad blockers and viewability) were three big challenges that digital
marketers face, but I had no idea that bots could mimic human behavior and put
items into shopping carts, for example. As the professor spoke, I realized that
recently my computer has been running loudly a fair amount of late, and that it
ate up my battery in about half the time it normally takes one day recently. I
wonder if it was a bot taking control of my computer (if that is the right way
to put that…?).
So, when I
ran across this article this morning from emarketer, it caught my eye:
The article
is based on data measured by Distil Networks, which is described as a “bot
detection and mitigation services provider.” It found that almost 40% of “bad
bots” are able to mimic human behavior and therefore show up as human traffic
on Google Analytics and similar tools. And, depending on the size of a site,
bad bots represent anywhere from 26–31% of a site’s total traffic (humans
represent anywhere from 15-60%, and good bots (the ones that are gathering
information used by organizations) comprise the remaining 9-54%. I was struck
not only by the percent of traffic that comes from bots overall, but also the
huge ranges in the percentages cited above. For a site that has 10 – 1,000 site
visits per day, another organization, Incapsula, found that only 15% of their
traffic was from humans! It didn’t say why, but I wonder if this could be
because larger sites have more money to spend on defensive measures? The graphic is pretty interesting:
And finally,
what I am not sure I understand is how these bots make money….is it that the site
publishers actually hire companies to inflate their traffic? I assume that’s
the case, but it is a pretty sad commentary on our society if that’s true.
No comments:
Post a Comment