Form Foibles - A Real Life Usability Lesson.

Way back in January of 2001, Peter Van Dijck performed a test called “Evolving Conventions” on PoorButHappy. In this experiment, Peter took a very "Yahoo-esque" Chinese Webpage and allowed people to submit their opinions about what each element of the webpage was – i.e. the sign in box, the search form, the directory headers, etc. You can see the experiment here.

Lately, due to all the subscribers that have joined the Makovision "Cutting Through The Crap" Newsletter, I have experienced a "déjà vu" of sorts. My subscription page has a little checkbox at the bottom, similar to the "Please send me more information by 3rd party companies" checkboxes prevalent in many sign-up forms.

The Guilty Culprit
View the actual form here.

When I look over the Makovision membership records I notice that a good number of new subscribers 'uncheck' this box during their initial signup. Arrrgh!

"Big deal," you say. "So they don't want to subscribe to your crummy newsletter." While I would tend to agree with this (well, except for maybe the 'crummy' part), the whole point of filling out the form is ONLY to subscribe to the newsletter. I merely placed the pre-checked checkbox there (supposedly) to confirm the subscription in the user’s mind - as if to say, “Yes I do want to be subscribed.”

Obviously I am not in touch with my users.

It turns out people are filling out the form in haste (remember, web users are a busy lot), and merely uncheck the box because they associate checkboxes at the bottom of the form as an invitation to spam. Additionally, another reason could be many of my subscribers are from foreign countries and may have a difficult time even understanding or reading English. Logically, they may not even attempt to read the copy, deduce the checkbox must be some invitation to spam, and uncheck it.

How many people? Well here are the basic numbers over the last 2 months:

This coincides with many basic marketing principles - both on and off-line. Marketing is a battle of perception, and quite obviously the perception has already been made in the minds of Internet users that checkboxes at the end of sign-up forms equal an evil only punishable by death, castration or some sadistic melding of the two.

Do not assume people know what you mean. Remember the user. Avoid anything users can’t stand – or anything even remotely related to things they can't stand (i.e. spam checkboxes, pop-up/under ads, etc.). Do not assume your users are carbon copies of you – if many have a hard time even understanding your native language, chances are they are going to skip the reading and I guess there’s more text to this paragraph?

Now I know what you are saying, "Wow Don. That’s really some rocket science you've got going here. ‘Avoid anything users can’t stand’? Oh great, another Jakob Nielsen wannabe who likes to state the obvious and charge exorbitant fees for stroking the unknowing and unwilling."

But I beg to differ.

Take Peter Van Dijck's Evolving Conventions test as an example. To be honest, way back in January I dismissed his experiment as more fluff than substance, but it really does hold weight. It is sad that people like me - full of themselves and considered pretty well informed on usability issues, can't even get a simple subscription form right. On the other hand, had I performed even the most basic user testing, I probably would have found out this checkbox thing was an issue early on.

Losing a quarter of your subscribers over a two-month period due to a simple form foible can be a very bad, evil thing in a company that lives and dies by these numbers. When money is your bottom line, don’t scrimp on the User Testing. While it took me months of data to look back in 20/20 hindsight, User Testing can give you near-perfect vision into the future success of your application.