You relentlessly simplify. You collect data and look for errors or opportunities for misunderstanding, then eliminate them. You tell people how to succeed with the app three times. You make the defaults close to success. You aggressively segment your users, giving the hard stuff only to those that can handle it. You provide an easy way to talk to you and pipe that straight to the dev team to automate or eliminate responses to the common issues.You implement game mechanics and award people features for learning or mastering other features.
This is not trivial but it also is not impossible. Trust me if my users can do it then yours can too. Help them succeed and take their money for it.
I think that the point is that there is not a single product at work here, but multiple products from different vendors. Facebook can't improve their interface to make users not type "facebook login" into Google as a way of accessing their site.
That is a failure of the imagination. They certainly could -- whether it is worth doing or not is another question, but hey, that is what God gave us A/B testing to figure out.
"Hey user, it looks like you came to us today from Google searching for [Facebook login]. Did you know that there is a better way? Type facebook.com into [blah blah blah]. Try it now and we'll give you 5 free credits for [without loss of generality: FarmVille]!"
"Great job! You should do that every time. If you do that to log into Facebook the next five days you use the service, we'll award you a Facebook Diploma and give you another 10 free credits for [without loss of generality: FarmVille]!"
On the back end, you show the above prompts to N% of your users who you detect coming to the login page from Google search results (this is trivial -- check the referer). You then compare any user metric you want for the "Was Shown Facebook Login Course" population and "Complete Facebook Login Course" population with the population at large. Kill the test if it hurts your metrics, deploy it sitewide if it helps them.
Incidentally, I was talking to a fellow Japan-based HNer the other day, and we both independently talked about IMVU. You know why IMVU is the coolest technology company I've ever heard of even though they make a product which I couldn't care less about? Because if this happened to IMVU, the most junior engineer in the company could code up the above behavior and deploy it live in less time than it took to read the blog post describing the problem, in full confidence that the production site would not break.
The possibilities of that development model are staggering. (Want to know how you can run rings around more established competitors? If you're learning at that speed and they're learning at the "release every six months" speed you will bury them.)
Answer: You can't, and no, you shouldn't try. You have to assume your users actually have the ability to get to your website before you can worry about what they do when they get there.
Sorry, but this whole thing is a whole new level of idiocy I had not previously encountered: Even my mother (whose sole computer skill is that she can get to amazon.com to buy gifts for people) would be able to determine whether or not she was at amazon after searching Google and clicking a random link.
Facebook users who have made this mistake are also confused by this. Clearly something is amiss--they know they aren't at the Facebook they desire--but they don't know where the mistake was made, or by whom.
Even the basic error is not egregiously stupid. Generally speaking, wouldn't you expect, from years of Google usage, that the first result for "facebook login" to be the Facebook login page? In fact, it is, if you ignore the news results. If.
Consider that: RWW is a top result for "facebook login" (not a "random link"), it does have multiple Facebook logos and the text "Sign in with Facebook" directly above the comment form. These things conspire to aid the user's confusion along with the most damning setter of expectations: the fact that it used to work.
You may understand that Google's results are not always perfect, that the Facebook logo appears on sites unrelated to Facebook, and that "Sign in with Facebook" does not have anything to do with using Facebook. But does Gladys? Evidently not. And why should she? Until yesterday, she didn't need to.
> Have them bookmark your site. Many of them will not understand bookmarks. Teach them to understand.
Isn't that really advocating that each and every website on the internet should teach their users basic computer skills?
I mean if these people had bought a magazine from the news stand, and it said "Popular Mechanics" instead of "Playboy" on it, do you think that they would just say to themselves, "I guess that Playboy changed their name, logo, and stopped showing pictures of naked women. Interesting. I'll have to tell me friends about this?"
A hypothetical engineer might be emotionally invested in "Everyone should know enough about computers to do basic things, and they should have learned that years ago, so if they can't do them it is their fault." That hypothetical engineer might bristle at being told "You should help them succeed, even if it means teaching them the basics."
A hypothetical business owner, told that there was an issue that was preventing his customers from paying him money, would probably say "Fix it."
Some days we have to pick whether we want to be engineers or business owners. For my part, I spent time yesterday teaching a PhD how to open PDFs.
> "Everyone should know enough about computers to do basic things, and they should have learned that years ago, so if they can't do them it is their fault."
I think that it's more along the lines of assuming that everyone on the road took a driver's test to be there. You don't think that people should have learned to drive years ago. People just feel that you should be able to assume a certain level of competency among your users as part and parcel of owning/using a computer. But as we see in the real world, you can never assume that people have a certain level of knowledge/logic ability. (i.e. Needing to put the "do not use top of ladder as a step" warning on ladders because there are enough people that don't realize that's dangerous)
Slightly akin to getting in a taxi, saying "Joe's place", ending up at in a house that is entirely unlike that of your friend Joe's, the furniture is different, etc and still being confused at why those in the house aren't happy to see you.
I'm early 30s and already prone to delegating rather than learning new technology, but I at least make the effort to read instructions, warnings and errors so I know what's going on.
We were talking about that at lunch and saying how we don't have the same mindset as "regular" people. For example, I just never ever click on ads. But many many people do. Similarly, I'm very reluctant to pay for virtual goods (in games mostly), especially when I know that what they provide is just a very simple programming trick (e.g. a game site selling "pens" to write in different colors in chatrooms)
I agree that it's actually difficult to get into this mindset. I tend to say that you can get money if you can "cater to the stupid". I thought it was kind of harsh but that Facebook login story proves me that it's not necessarily.
Depends what virtual good your paying for, more content in a game I'm spending a lot of time playing would be worth it, although I don't see myself liking any of the flash app games on that level.
Google ads can be useful from time to time, random ads not really, I think most of the random ads I click on are non purchase interesting looking things or ads for physical stores I know because they have a decent deal in the ad.
Educators in the US are mostly boorishly uneducated.
A Masters in Teaching is a professional degree (it exists to get the holder an automatic raise), and even by those low standards it's at the very bottom of the barrel, even lower than the average MBA. A doctorate in education is closer to a normal masters degree.
"Educated" is the past tense. Some people just stop learning after a while, even though there were capable of it in the past.
People that have education and continue to get it are the people that can "avoid these problems"... but we don't really have a good word for "educated and able to learn", so the OP just said "educated".
Perhaps they have managed to get a Master's without actually becoming "educated." It's not like it's not possible—you can become an expert in a very deep-yet-narrow field without ever having to learn how to think and problem-solve in the general case (which is what a liberal arts degree was supposed to be for.)
I can't create a single sentence in correctly written Chinese. Or Russian, or even French for that matter -- I can cough up some random French words on paper, but they will likely be misspelled and "just plain incomprehensible."
What "type" of person am I?
Written communication is a stunt that is mastered by a relative handful of people, and polite, formal written communication by even fewer. The others communicate by more traditional and universal means: Talk, music, dance, art, games, cooking, commerce.
These people are primarily native English speakers. Of course errors are excusable when mastering a second language, but I don't think anyone seriously believes that any significant number of those people are non-native English speakers.
This is pretty much how I use google, and yet I'm a coder. We seem to be assuming that these are only non-technical users. I find the quickest way to get to what I want is usually through a google search rather than visiting a site and browsing around within it trying to find the feature or content I want. Basically I've adopted the search interface as highly more efficient than most user interfaces within most websites.
However I do actually read my search results and refine my criteria if necessary rather than just clicking on the first link.
Look at the interfaces exposed by any general public machine/gadget outside of computing, and the layers of abstraction that the consumer needs to grasp.
Automobiles are one of the rare gadgets that expose an internal system layer (the engine) and right there you clearly see a divide between the users who can only interact with the general purpose (simple wheel, pedals, key, gear shift) interface, and the automotive geek ("grease monkey") who also has a clue what's going on under the hood.
As soon as any sort of symbolic interaction is required (remote/vcr programming) you again lose a whole chunk of the population that simply can not get it.
A "consumer product" has to be an appliance, not a general purpose computing machine, hosting a general purpose operating system, hosting a general purpose application (web2.0 browser), hosting yet more general applications.
The fact that the "farmville" crowd even manages to get to interact on the web (and leave hysterical comments on RWW) is by itself a major accomplishment of the geek set! ;)