Lies, Damned Lies & Statistics
(or The Importance of Knowing What The Hell You’re Talking About).
Earlier today, Vicky Brock posted a graph to Flickr labelled “ICO website traffic impact of cookie opt in“. She had compiled the graph from the results of a Freedom of Information request to the Information Commissioner’s Office, which had yielded their Google Analytics data.
Before long, tech blogs were writing stories based on the data. Chinwag reckons it’s “cookiepocalypse“, while TechCrunch asks whether you “Want a 90% drop in your site visitors?“. To the casual reader, it would appear that implementing an explicit cookie opt-in, as the ICO have done, will result in 90% fewer people visiting your website. However, this is not the case at all.
The graph shows the number of visitors as measured by Google Analytics. Anyone who, like me, was willing to spend all of about 60 seconds researching how Google Analytics works, would have discovered that Google Analytics uses cookies to track website visitors. If a visitor’s web browser doesn’t accept a cookie from Google, their visit and page views won’t be recorded by Google Analytics.
Before the ICO implemented the cookie opt-in on their website, you had to explicitly configure your browser to reject cookies (whether in general or from Google specifically). The vast majority of web users probably don’t even know what a cookie is or realise that their web browser downloads cookies from almost every website they visit, let alone know how to restrict their browser’s appetite for cookies. Hence, most people accept cookies unwittingly and, as a result, Google Analytics can track most people on the Internet.
By introducing a cookie opt-in, the ICO turned those numbers on their head. Instead of automatically trying to stuff a cookie down the throat of every person who visits their website, the ICO will now only give a cookie to people who tick a box and click a button to indicate that they’re happy to receive a cookie. The opt-in box is at the top of the page, isn’t unnecessarily large and isn’t obnoxiously sticky (i.e. it disappears off the top of the screen as soon as you scroll down the page). Most visitors to the ICO website don’t need to accept a cookie in order to find whatever information they were looking for.
So guess what? Instead of circa 99% people receiving cookies without realising it, only about 10% of people went to the trouble of ticking the box and clicking the “Accept” button.
That is what is reflected in the data that Vicky Brock obtained from the ICO and the graph she posted.
In other words, a whole lot of people (some of whom are clearly not qualified to comment on a topic like this), are making a mountain out of a very small molehill, jumping to conclusions and prophesising doom and gloom. There are a slew of appropriate quotes here: “Those who know, speak. Those who speak, don’t know.” and “Empty vessels make the most noise” spring immediately to mind.
This is a common problem in the Internet/tech industry. Like any fast-growing sector, it attracts self-professed “experts” who basically know more about courting publicity and exposure than they do about the actual subject matter.
The moral of the story? Caveat audiens.
Another great post! The cookie-scare has gone way overboard.
Nils Pihl Bohlin
September 6, 2011 at 2:23 am