AIAIO: Our Blog

Ai braves the NYC blizzard to grab some Shake Shack, and even got some free hot chocolate!

Ai had a great week at Internet Retailer’s Web Design and Usability Conference in Orlando. We’re still happily processing the many people, companies and ideas we encountered.
As in the past, Ai’s crack user experience team gave complimentary design consultations to conference attendees, ranging from major department stores to mom-and-pop businesses. We set up camp and enjoyed two busy days of consults.

More notably, Ai took a booth on the expo floor, a first for our company. We met a lot of interesting folks and learned plenty, which we’ll put to good use at IRCE in Chicago in June.
We also learned a lot from the conference itself, particularly from Lauren Freedman’s annual post-holiday metrics roundup. Among the news: 40% of the top 100 ecommerce sites held Gilt Groupe-style flash sales last fall; and of the 60% of sites that have instituted social networking links, fully (only?) 85% of them point to Facebook.

This author enjoyed a moment in the spotlight–two, actually–doing a pair of live website critique sessions on the main conference stage. Everyone from Big Al’s Aquariums to B&H Photo Video got in on the action.

But our proudest moment came Wednesday morning, when Jack Love, publisher of Internet Retailer, took the stage to show the room Ai’s designs for the new Internet Retailer website. We can share the news now: Alexander Interactive, creator of many sites in the Internet Retailer Top 500 Guide, is now working on IR itself. We loved gauging the crowd’s reaction, which was similar to our own, and are diligently working on getting the site ready for a spring launch.

If you saw us at IR and want to follow up, drop us a line. See you in Chicago.

The Internet Retailer Web Design 2010 conference confirmed one of my long-held assumptions about the online retail industry: most ecommerce sites underutilize their metrics.

This is actually an understatement. Most companies, particularly small businesses with less than $10 million in revenue, barely use metrics at all. And they’re missing out on huge business opportunities as a result.

By metrics, I refer to the page-level data about user behavior that can identify problems and unseen possibilities to improve an ecommerce (or any) website. This isn’t about sku-level data and revenue per order; on a base level, every seller knows how much he’s selling. This is about knowing how and why those items sell, and how a website is affecting performance.

Time after time at IRWD I asked a retailer about site performance, and each time I got blank stares or embarrassed grins. “We actually don’t have any analytics installed,” some people told me. Many others routinely gave a variation on, “Yeah, we have to take a look at it.” Of all the people I met, I only spoke with one business that was able to recite to me some of its important data points and the efforts being made to improve on them.

Data comprehension may be the single biggest factor in getting incremental growth out of a site. By looking at analytics, a retailer can know what pages are underperforming, where fallout occurs, and how processes and displays can be improved. Simple, free or inexpensive tools can report everything from visits before purchase to the location and timing of mouse clicks, all of which are part of the smart retailer’s arsenal.

But instead of building on these smarts, I got interactions like these:

“Hi, David, I was wondering if you could look at my website. I have two sites selling similar merchandise, and this one has only a third of the sales of the other one. This is the redesign I’m working on for the worse site. Do you think it will help?”

“Well, to start, have you looked at your analytics?”

“No, we have a Google account, but I just know that this site isn’t performing.”

“You have a site that’s up and running and doing well. If you crack open your Google Analytics you can compare the data side by side and see where the lesser site is underperforming. That way you’ll know what to attack. You’re actually in a great position to fix it if you do that.”

“But do you think this new layout is going to help?”

I don’t know if that redesign will work, but I do know that the owner of the site is ignoring his own benchmarks. If he looked, he’d be a lot closer to the answer.

Why dive into a redesign without identifying targets for improvement? “Fix my conversion” is not a useful goal. “Make changes to the product page, which has a high fallout rate, to improve adds to cart while encouraging further exploration” is an actionable one. That actionable goal can be identified quickly and easily, but so few businesses are doing it.

Many, many questions raised at IRWD can be answered in-house with a thorough study of site analytics. Companies that ignore it are almost literally leaving money on the table.

Ai is pleased to announce that our very own Mayor Jack Reynolds has been nominated as a finalist in the Amusement category of the 13th Annual SxSW web Awards. The website, www.alexanderinteractive.com/jack, would never have come to fruition had it not been for everyone’s favorite dog on the Internet. There is a wide array of Mayor Jack’s video collection, the largest and most eclectic photo gallery of anyone online (man or beast), and a selection of buddy icons to download. Mayor Jack is quite the social butterfly, so there are also plenty of options for you to connect with him.

Mayor Jack Reynolds had this to say about the award nomination: “…”

Actually, he was sleeping. But it is safe to assume that if he was awake, he would have been grateful that he could inspire a website and if we all know him well enough, he is both humbled and excited to be considered for the award.

The SxSW Web Award Nominees can be viewed at sxsw.com/interactive/webawards/finalists, and the voting takes place online at sxsw.com/peoples_choice.

A minivan-ful of Ai peeps landed in Orlando yesterday for the Internet Retailer Web Design conference. We’re speaking, consulting, and showcasing on the expo floor, as well as schmoozing and playing far too little golf.

Keep an eye on the aiaio Twitter feed for updates throughout the conference.

I booked a flight to Austin for SXSW Interactive on Friday. Thanks to delays in planning and confirming my travel, I paid handsomely for the privilege: $674 for well-timed nonstop flights on JetBlue.

It didn’t have to be so pricey. For $419, I could have flown on Continental Airlines instead. But Continental’s booking system so frustrated me that I spent an extra $250 to fly another airline.

Some background: those who know me personally are aware that I don’t much care for Continental. But I’m also not one to splurge needlessly, so when I found out Continental’s EWR-AUS flight was a third cheaper than JetBlue’s JFK-AUS route–at similar times, on bigger planes–I figured I’d give Continental another shot.

I used Continental’s online reservations system to select my flights, then proceeded to the seat selector, which showed each flight at around 85% full. The return flight’s seat map (click to zoom):

The situation was the same each way. The flight had 15 seats available. Continental had declared all of them Premium Seating, even several middle seats, which meant I couldn’t sit in them. But the plane had no other seats available, which meant I’d be booking without a seat assignment.

More background: I’ve traveled enough to know that the guy with no seat assignment is the first to get bumped in case of overbooking. Continental had seats but wasn’t offering them to me. Worse, Continental didn’t have an alternative, just blocked, empty seats.

I understood Continental’s desire to hold good seats for its good customers. I’ve had preferred status on and off in the past and I respect the privileges that come with frequent patronage. But with the rest of coach filled, I couldn’t figure out why Continental wouldn’t give me an empty seat and confirm my travel. Besides, the map confused me: is seat 7B really a top choice of elite frequent fliers?

So I called customer service for help. The friendly Southern woman who took my call confirmed what I was seeing: yes, there are premium seats available; no, you can’t have them. I asked if I could pay extra to reserve those seats: no. I asked if I could get a seat assignment, any seat assignment, so I knew I would make it on the plane: no.

I eventually gave up my attempts to cajole customer service into helping me, and after a few hours of deliberation, I took my business elsewhere.

The user-experience takeaways here are twofold. One is pure information design: don’t share information that’s not actionable. All Continental achieved with the seating chart above was to drive me crazy, showing me that it had seats–some of them rather mediocre seats I’d typically avoid–that I couldn’t reserve. Had they just shown them as unavailable, by having me log in with my (non-elite) OnePass account before selecting seats, I’d have been far less frustrated.

The other, of course, goes to the heart of customer service: sell your goods to shoppers who desire them. Continental lost my business because corporate policy dictates that the booking system has to be ready to accommodate a dozen Elite-status fliers who might want to fly between Newark and Austin on a pair of weekday flights that arrive close to midnight. Why not acknowledge the demand curve and give a paying customer the seat assignment he needs to book his flight?

Even better, why not implement a policy that generates both revenue and customer satisfaction? Many airlines charge for preferred seating. Continental could have levied a $100 fee on me for its premium seats, and I’d probably have paid it, because I’d still have saved money over my JetBlue option.

Instead, I’m back on JetBlue, where I’m willingly overpaying for peace of mind and a guaranteed seat. Oh, and satellite TV in a leather seat with good snacks. Happy jetting.

I’ve been reading and thinking a lot about the iPad (how do you not?) since posting my initial, underwhelming reaction to its launch announcement last week. And, suffice to say, I’ve changed my opinion rather drastically.

The iPad will be, for many, greater than the sum of its parts.

Because on a checklist, the iPad disappoints. But then, this happens a lot. Consider the first iPhone: no physical keyboard, no picture messaging, no video, etc. But the device quickly overcame those detractions.

What the iPhone had, and what the iPad probably has, is totality. Apple’s genius in start-to-finish user experience design is well documented, and the iPad will be no exception. Interactions large and small, short and long, will all be easy to execute and fun to try. And it won’t take long before people yearn for one.

I noted last week that it’s likely a perfect complement for someone with a desktop computer: one device for heavy lifting, one for casual use. And the more I consider it, the more I think that’s the likely scenario.

This gadget is also a harbinger of the future. Apple is banking on people working more in the cloud, using shared and networked services, and interacting in new ways. It’s already happening; Ai, for one, despite griping about its flaws, uses Google Docs extensively. I know my own life has shifted from POP email and file hierarchies to IMAP and file search. Throw in a few multimedia subscriptions and the entire paradigm could shift for good.

Plenty of people have said this similarly and more eloquently than me; I particularly like Mike Monteiro’s observation, that the iPad is the cinema’s device of the future. So add me to the list of believers, with a dollop of humility.

The semantic web is an ever-evolving development to define the meaning of content viewed on the web. It is the responsibility of an HTML developer to build web pages with a sharp understanding of how the front-end markup should be structured. With the upcoming release of HTML5, new tags will be introduced to the language that will enable more fluid communication between the web and the people and machines that use it.

The common practice of laying out HTML markup these days is the use of the <div> tag, using id or class attributes to block out content. New tags included in HTML5 will display a much more readable layout when defining areas of a web page. Some of the important tags in this group are <section>, <aside>, <header>, <footer>, and <nav>. It is likely that most of these tags clearly state what they are doing, but each still deserves it’s own explanation:

• <section>

  This tag will be used to define generic content within a document. While id and class attributes will still be needed, this will still lay out a cleaner mode of blocking content than the now-standard <div> tag, thought the latter tag will not be deprecated and can still be used.

• <aside>

  Sidebar content will be separated using this tag to show content that is “to the side” of main content areas. Generally this content is partially related to the primary focus of a given web page.

• <header> and <footer>

  The topmost and bottommost blocks of content are often referred to as the “header” and “footer.” Naturally, HTML5 introduces tags that clearly define this. It is likely that id and class attributes will no longer be needed with the use of the tags, unless a developer opts to use the <header> tag inside of a <section> tag.

• <nav>

  Navigation menus will have their own tag in the markup, however it is uncommon for modern websites to have less than three nav menus on a single page, so id and class attributes would still be needed.

Below is a screenshot of how this code might look. The new tags form a much more readable structure to the markup of the page. It creates a much cleaner view and gives the viewer a better of understanding of what is being displayed.

View a live example (it may not be the prettiest site, but it’s more for the markup than anything)

One of the most beautiful aspects of the above examples is the HTML Document Type Declaration, or DOCTYPE. It is the absolute first statement in every single web page, and is used to trigger standards mode in a browser and then determines what iteration of the language is in use. In earlier versions, there was specific data required as part of the declaration, which may look something like this:

<!DOCTYPE html PUBLIC “-//W3C//DTD XHTML 1.0 Strict//EN” “http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd”>

Earlier versions of HTML were based on SGML, Standard Generalized Markup Language, but HTML5 is not and therefore has no need for the extra junk in the DTD. Here is the new one:

<!DOCTYPE html>

As previously stated – beautiful.

There are a few other new tags that are still worthy of mention: <article> to block content for blog entries or news articles, <dialog> used for marking up a conversation, and <figure> to be used when associating a caption with embedded content (audio or video).

It is debatable whether these new standards should be used, considering that full cross-browser support for HTML5 is not there yet. Firefox, Safari, Chrome, Opera, and a few smaller browsers will support most of these new tags, but older versions of Internet Explorer will not. Allegedly, there is a way to use JavaScript to force Internet Explorer to acknowledge that the new tags exist, but if a user has JavaScript disabled in their browser, the page will not render properly. Given that IE6 and IE7 still have a large-enough share on the browser market, this hack method does not seem sound enough to use in common markup yet.

The direction the semantic web is going will no doubt enhance the overall user experience of the web, even if the user is not consciously aware of it. This markup overhaul will create a better environment to display content and has the potential to make web pages even just slightly faster and more scalable. These are all good things.

GrokDotCom has a nice writeup of a too-eager lead-generation strategy on rent.com. Try too hard, ask for too much too soon, and instead of securing leads, a site just scares people away.

I like the post’s takeaway, which summarizes things well: “Remember, it’s not about you or your sales process. Your visitors are volunteers in the process and are coming to your site with motivations and intent.”

Whether on a home page, a product page, or deep into checkout, visitors have their own reasons for being on a website. The site can only do so much to steer those users into a set action. Don’t try so hard! Give people accessible options and let them do what works for them.

Ai is pleased to be mentioned not once but twice in this month’s Internet Retailer. (We’re in the print magazine as well as the website.)

The cover story, The dollars are in the details, quotes this author as well as other industry experts on how to nail the details in UXD. Ai’s angle, as always, is to combine analytical research with active listening and observation of users’ needs and behaviors.

More expansively, Alex Schmelkin and Josh Levine are quoted at length in a profile on customer-centric web design. Our mantra, as always: understand motivations, empower users, and test, test, test!

In addition to being mentioned in IR, Team Ai is also going to IR: in two weeks we’ll be in sunny Orlando for the Internet Retailer Web Design 2010 conference. Ai has some serious coverage this year, which we’ll be detailing in a news item shortly but includes:

• Two days of private design consultations, run by Josh Levine and Ed Samour
• Two live on-the-fly site critique presentations, featuring yours truly
• Ai’s first expo floor booth; stop by and say hi!

With numerous clients in the Internet Retailer 500, we’re excited to be such active participants this winter.