Search Authors: Michael Bushong, Shelly Palmer, Yakov Fain, Kevin Benedict, Dmitry Sotnikov

Blog Feed Post

Why Are CAPTCHAS So Awful?


Yes, the dreaded CAPTCHA. CAPTCHA, or “Completely Automated Public Turing test to tell Computers and Humans Apart” (yes, someone actually thought up that phrase), has become commonplace across the Internet. From trying to order tickets on TicketMaster.com or Stubhub.com, to ordering a new pair of boots from your favorite Winter boot supplier (hello Sorel!), to proving your a human when trying to login to your locked online bank account, the CAPTCHA is commonplace across the Internet. This said, the CAPTCHA is awful.

To understand why the CAPTCHA is awful, we need to discuss the original purpose of CAPTCHA and the downfalls inherent in the commonplace security system.

CAPTCHA is a Selective Blockade

CAPTCHA, much like Akismet spam filter, was designed to eliminate spam bots from utilizing a site to create fake accounts/orders. The idea was pretty simple – create an image unreadable by bots to ensure bots do not create any fake accounts/conduct any fake ordering. The original intention was excellent, however in practice, the intention turned into a selective blockade for both the sight impaired and the seeing. Studies of CAPTCHA have revealed time and time again out of group of 100 people, only 30 could correctly identify the CAPTCHA in question. When the study expanded to a greater sampling size, it showed less and less agreement on what the CAPTCHA was actually stating. Point in case, tell us what the first word in this Facebook CAPTCHA is:

Terrible CAPTCHA

Clearly the second word is “concerns”. Does anyone have a guess to what the first word is? Is it even possible to identify that word? If this CAPTCHA confusion proved to be an irregularity, the issue would be mute. However the issue isn’t an irregularity. No, CAPTCHAS which are down right impossible to read are normal across the Internet. If you are looking to purchase a product online, CAPTCHAS are frustrating enough to spur cart abandonment (more on cart abandonment in a moment).

The other major issue with CAPTCHA security measures is how they discriminate against anyone who is sight impaired. For the sight impaired, utilizing a CAPTCHA is a no go. To rectify this, some companies and ecommerce platforms have opted for an audible CAPTCHA. While the audible CAPTCHA automatically eliminates anyone who is hearing impaired, it also has proven harder to decipher than the normal CAPTCHA. Don’t believe us? Just take a listen to this very commonplace audible Google CAPTCHA:

We know. What the hell was that? If you can tell us, either you have a range of hearing which no human being on the planet has or you are David Blaine. Either way, this type of audible CAPTCHA fail should not be taking place and yet, across many platforms, audible CAPTCHA techniques prove less reliable and more annoying than onscreen CAPTCHAS.

CAPTHCAS Are Killing Your Conversion Rates

In August, the search engine ranking firm, Moz, posted an excellent blog post concerning how CAPTCHAS are killing your company conversion rate. The reason was simple: for clients looking to buy a service, why in the world would you place an unneeded hurdle in their path? The CAPTCHA proves to be an impediment for clients looking to purchase a service, so much so that cart abandonment is common.

As noted in the Moz article,

“…according to a study carried out by Stanford University into the use of CAPTCHA by humans. Yet, by testing more than 1,100 people, gathering 11,800 completed surveys, and studying 14,000,000 samples from a week’s worth of data from eBay, they revealed just how difficult CAPTCHA has become for humans. The study showed that, on average:”
  1. Visual CAPTCHAs take 9.8 seconds to complete
  2. Audio CAPTCHAs take much longer (28.4 seconds) to hear and solve
  3. Audio CAPTCHA has a 50% give-up rate
  4. Only 71% of the time will 3 users agree on the translation of a CAPTCHA
  5. Only 31.2% of the time will 3 users agree on the translation of an audio CAPTCHA
With around 1% of the audience currently using audio CAPTCHA, this is potentially a huge market to lose.”

CAPTCHAS are killing your conversion rate. 9.8 seconds to complete for visual CAPTCHAS and 28.4 seconds for audible causing a 50% failure (cart abandonment) rate.

All this said, the question needs to be asked: how can we fix CAPTCHA? What are some viable CAPTCHA alternatives?


Multiple avenues have been proposed to fix CAPTCHA. From using the Honeypot Method (add CSS fields invisible to humans, auto-filled by bots), to using mathematics logic questions (What is 7 + 23?), to utilizing a simple check box to prove you are human, to playing interactive games or drawing shapes, alternatives to CAPTCHA exist. The question is, which one will work the best to not only solve the issue of CAPTCHA annoyance and cart abandonment but will also solve the issue for the hearing and sight impaired?

The answer could be simple. Instead of automatically assuming every customer and potential client is out to harm the sales process and commit fraud, it might be a good idea not to spring a CAPTCHA until an actual problem arises. As noted by Solar VPS Lead Programmer Matt McCaffrey:

“not to preemptively assume there is going to be such an unmanageable problem, that every good person must be made to suffer..”

He makes a good point. Only use a CAPTCHA when a real problem arises. Don’t use it to piss people off.

Whatever the solution, we can confidently say regardless of the fix, the current state of CAPTCHA is down right awful.

It needs fixing.

Now, please, if you don’t mind, decipher this…if you can.

Terrible CAPTCHA

Read the original blog entry...

More Stories By Solar VPS

Solar VPS lives the Parallels "Optimized Computing" vision. It has created a virtual infrastructure from client offerings, data facilities and management offices. Solar VPS works very closely with Parallels to provide the highest possible service and support to customers.