Unbelievably straightforward, yet unbelievably effective – CAPTCHA is a means of determining which visitors to a site are genuine users and which are automated programmes. By utilising the differences between human and machine thinking, it acts like a puzzle lock on websites – one that software simply cannot understand how to grasp.
What is Captcha?
Captcha stands for ‘Completely Automated Public Turing test to tell Computers and Humans Apart’. In plain English, it’s a randomly generated test that humans find exceptionally easy but computers find neigh-on impossible. Found on all kinds of sites, it was developed with the aim of preventing spam and normally takes the form of a quick maths question or a letter comprehension task, although more recent iterations ask its user to pick out objects from a photograph.
Why was it developed?
Before its introduction, it was fairly straightforward to play the system. Spambots were regularly deployed to automatically post comment spam, drive false traffic to a site or resubmit a form thousands of times. It meant that user experience deteriorated and online ranking systems could be rendered virtually meaningless. The challenge lay in a finding a solution that could filter the spam while reducing its impact on the user. Enter captcha.
How does it work?
A computer may be able to decipher the word “Yellowball” but were it written as “Y3\\0w8A\\” it’s almost certain that its true meaning would go under the radar. It’s this fundamental difference in human and computer perception that forms the basis for all captcha tests. When it comes to pure logic, artificial intelligence has surpassed even the most intelligent of humans; however, when it comes to picking up on context and inferences humans are still the top dog.
A calculator now holds more mathematical processing power than the Mensa qualified genius who holds it. Ask it to tell you which number takes up the most physical space though, and you’d be better off writing the question down on a slip of paper and leaving it underneath a rather inviting rock. The key to a captcha test lies in the fact it simultaneously requires attributes only possessed by humans. When deciphering distorted text, humans can recognise and filter out random lines that strikethrough the page, an inherent result of our social upbringing. Likewise, they can differentiate two overlapping characters by recognising the fact they must form part of a wider word.
Reinventing the Captcha
The increasing intelligence of AI means that traditional text-based captcha tests are no longer as foolproof as they once were, to counter this reCAPTCHA was developed in 2008. It works on a similar principle but uses images as the source material. Originally it killed two birds with one stone, by asking the user to enter scanned text from Google’s archives. One word acted as a control whilst the other was added to Google’s extensive digital library – helping to record the contents of historical books and newspapers for future generations.
Of course, no matter how small the inconvenience, people are naturally averse to any form of testing. It’s with this in mind that Google introduced the latest generation of captcha in 2014, rather confusingly termed the ‘no Captcha reCaptcha’. At its core, it does exactly the same thing as its predecessors, with one major advantage. It’s invisible.
The exact way in which it works is a secret, closely guarded by google. However, it’s thought that the way in which a human naturally interacts with a page is different to that of a bot; the way the mouse moves, the length of time it hovers before clicking, all of these are subtle hints that a user is who they purport to be. Occasionally a particularly robotic looking human may fail the test – triggering a confirmation captcha. These cleverly feedback into improving google’s services, for example, identifying door numbers from a Streetview image helps to improve google’s maps service, all of which ultimately aids its accuracy and user experience.
How to make the most of Captcha
Captcha means that it’s possible to instantly safeguard your site from an array of spam, keeping it clean and relevant to the user. However, traditional captcha has been shown to impact the conversion rate, with users taking it as a personal affront. To think of it one way, people walking into an art gallery aren’t automatically thought to be there to graffiti the place, so why are visitors to a site similarly assumed to be spam bots?
Fortunately, the no Captcha reCaptcha does away with the negatives, leaving a largely seamless user experience. To implement it on pages of your site visit Google’s official developer guide, where you’ll be advised what code is needed to run it on desktop and Android. Luckily it’s relatively quick and it’s easy to do. Well, unless you’re a robot that is.