Nielsen concludes “Ideally, the next step is to user test the redesign. You can do this with paper prototyping, so you don't have to fully implement a new app . . .”

Rather than that, it's actually quite easy (and overlooked IMO) to implement a meaningful A/B framework within iOS by setting up your experiments as live web pages that the app accesses.

The key reason NOT to do this is latency, so make sure you're on the fastest box possible and consider pre-caching the experiment images via SDWebImage or similar.

In our experience any loss you have around latency will be made up in spades by having an engine that lets you test your funnel in a real world environment.

ref for image pre-loading -

https://github.com/rs/SDWebImage

http://stackoverflow.com/questions/4617424/improve-uitablevi...

http://www.iphonedevsdk.com/forum/iphone-sdk-tutorials/13315...

Does this rely on having internet access available on the iDevice? Or could you bundle default screens in your app, which only update if internet access is available?

yes, you can cache pages locally, and update when internet access is available

I wonder how popular A/B testing is on iOS.

I agree with you though, I think most people avoid it just because of the very nature of the app store..

See that isn't quite right, while the app is free to subscribers to the physical newspaper, those of us who use the online web access don't get access to the iPhone, only the iPad. That's why I'm unhappy. Dudes mockups still leave us online users pissed.

Thank you! Someone is paying attention.

This article makes it look like it's free to online subscribers (if you can figure it out).

But that's not the case at all. It is NOT free to online subscribers. Therefore it's got bad design, cryptic pricing, and the are double charging. All of which warrant two star reviews.

A cynical person might say that the design was intentional, to vacuum up a few extra subscription fees. Especially when the actual login button is shoved all the way at the bottom of the screen and significantly smaller (due to a lack of a background shape) than the other two buttons. Buttons don't shrink on their own accord, especially when it comes to interface design. Whoever wrote the app had to have the stakeholders sit down and give input on the look and feel - I doubt they'd miss something like that.

Statements concerning the need to test aside, "almost unusable" is very likely a gross overstatement. To state http://useit.com is almost unusable is equivalent to saying, at the very least, that it is hard to use. It probably could be easier to use (what isn't?), but hard to use it is not.

You can argue that the site is not the most aesthetically pleasing. There may have been a time when this could have been changed, however as the site and Jakob Nielsen have been well known for many years, the visual appearance serves strong branding purposes: it's very obviously useit.com, it's very recognisable, even standing well back from your screen. Jakob has supported this notion as the reason not to change.

Almost unusable?

Your comment is almost unreadable. How can you possibly say that site is almost unusable?! It's so simple. Search works excellently. Site layout is clear, obvious. Fonts are easy to read.

Hell, it's probably a lot more usable than HN.

No, you're confusing usability with being pretty.

No.

If one looks at useit.com homepage, what is "Alertbox" to the left supposed to mean? Are they blog posts or articles? If articles, how do they differ from "Reports" below? And in the News section to the right, both the news items and "as seen in the media" are mixed together.

Jakob has been writing Alertbox columns since 1995, clicking Alertbox takes you to an index of all the columns. Reports are long format studies of usability issues that are available for purchase. Click any report title to see a summary, TOC and purchasing options. There is a horizontal rule separating media mentions/interviews and the lastest on his events. I find nothing about this confusing, it's quite usable. It's not pretty though.

To be fair, what you're highlighting is that useit.com caters better for regular visitors than for newcomers who aren't familiar with its terminology, which is a genuine usability problem. Your visitors spend most of their time on other people's sites, remember? :-)

Is Alertbox really such a cryptic term? I doubt that very much.

It's misleading if you don't know what it refers to - it makes one think it pertains to something that they need to be aware of, perhaps even something dangerous... maybe a security issue or something like that.

I guess Nielsen is aware of this issue. “Alertbox” is the only headline on his start page (save for the actual title of the website) that has a subtitle: “Jakob's column on Web usability” – a quite succinct description of what the Alertbox actually is.

Calling it Alertbox is certainly more about branding than anything else. (People don’t read on the web; the subtitle is not going to help much, right?) Nielsen has been writing his column since 1995, renaming it “Jakob’s blog” or “Jakob’s column” or something like that at this point in time would be extremely stupid.

Everyone who was somewhat interested in usability during the last decade knows about the Alertbox (if only because they were annoyed by it). It’s a brand. All the noobs (who this website isn’t really targeting anyway) get the subtitle.

It was noted by Amy Schade during the 2008 NNG Usability Week conference that "Alertbox" would not be a wise choice if selected now but that the branding is too strong to support an easy change.

"Being pretty" is a must for great usability. I am not talking about decoration which is superficial, but the real prettiness which comes from having clear layout, visual hierarchy, well thought-out typography, etc. To borrow from Steve Jobs: design is how it works. There is not much to "work" on useit.com, but there definitely is a room for improvement (layout, typography, focus, etc), now it is just two columns with bunch of links. See for example http://alistapart.com/ — for me it is still has the best design for content-oriented site. It is very pretty and at the same time all that prettiness has a function.

I see clear hierarchy, descriptive titles, good placement of standard elements (search is at the top right, contact info is at the bottom), appropriate font sizes and good contrast.

I really dislike the visual design but that doesn’t make the website unusable. It makes me less happy when using the website but I don’t think it impedes my ability to go where I want to go and to do so quickly.

Did you read it? This isn't a "your site isn't usable to the 1% of users with motor disabilities" Nielsen article.

WSJ spent money to create an app that, because of one poor piece of text, made existing, paying customers dislike them.

And not for a real reason, like charging them twice, but a misunderstanding caused by bad copy and never testing the app with real subscribers.

First, how Nielsen came to conclusion that the reason for the app's bad rating is its startup screen design is beyond me. Read the comment of cageyjames. That could be the reason.

Second, usability isn't a science. So it's vulnerable to criticisms.

> Second, usability isn't a science.

You seem to be using a different definition of "usability" to most of us throughout this discussion. To people like Nielsen (who certainly popularised and arguably originated the term) usability is like a science: it is about measuring how people behave in practice and improving your product in light of that hard data. Once may draw inferences from that data, if patterns start to emerge where some practices are on balance superior to others, but just as in science, such inferences are falsifiable and only as good as the latest data set.

It is more than a little ironic that useit.com itself has become less usable over time, according to many of the very factors that Nielsen himself has commented on in the past...