Babysitting the interface: The myth of self-service

As we prepared for another road trip and the attendant joy of air travel in America in 2010, I was musing about the common sight of a human being helping passengers use the so-called "self-service" kiosk. It was in that frame of mind that I went to Facebook; I had to laugh when I saw this screen with the instruction: DO NOT CLICK THE "Go to Application" BUTTON ON LEFT!

Human intervention with software, Web pages, and kiosks is a regular occurrence across corporate America. You see it first-hand at airport counters, supermarkets, a co-worker's desk. We see it inside call centers and controlled environments. How much does this cost business, education, and nonprofits every day?

Well into the year 2010, we are still seeing an utter failure to budget - or even account for - the design of information structure, meaning: Where will this digital media experience begin and end? How does the user know where he is in the morass of content? How do you know whether you've seen and experienced all the content? How does he get to related content? Did anyone design or plan this experience? What are the opportunity costs related to this basic failure of planning?

In 2000, seven years into the commercial Web, a failure to plan and structure digital media interaction was understandable. Skeptics were still expecting (or hoping) that the Web would die. In 2010, the failure to plan and structure digital media is unforgivable.

If your digital media project lacks a budget item for information design (information architecture, task sequence design, user interface design), it is destined to under-perform and disappoint. Web sites and applications, software applications, kiosks, intranets, instrumentation interfaces - any digital implementation must include planning and architecture beyond simple functionality and appearance.

The Self-Service and Kiosk Association begins to address the usability issue in a blog post by Stephen Kendig, but the industry is just beginning to scratch the surface. If we use the Web as an object lesson, bad information design proliferates more quickly than good design. We will be living in a dystopian reality if the same problem propagates across self-service in a world that seems determined to eliminate human-powered customer service. The current economic woes will only exacerbate the problem.

If we're going to eliminate the human from the equation at all the transaction points of daily life, the digital interface must be bulletproof. Else we are destined for the worst of both worlds - unhappy customers, unhappy workers. Babysitting an interface is no one's career goal.

Who exactly is taking my money? Poor usability = reduced online sales

It's ironic that online check-out - the point in ecommerce at which users are most sensitive - is the point where many Web sites fail to provide the brand consistency that users need to feel confident about entering personal information and credit card numbers.

Abrupt interface changes (from the main Web site to the ecommerce application) are the most common - and most destructive - usability problems on ecommerce sites. A startling new interface with different branding, combined with a new URL, will give pause to any user, regardless of technical literacy. They wonder: Am I on a different site? Is the privacy policy different? Who exactly is taking my money?

When someone decides to purchase from your Web site it's because they trust you and your brand. Our own research and usability testing has shown scores of users halting and not completing transactions when they encounter an unfamiliar screen at the point of interaction.

What caused this problem? Initially, many Web sites did not have ecommerce functionality, so it was bolted on as an afterthought. Third-party applications were used as the ecom stopgap. Brands were not allowed to customize these interfaces. The result: A bumpy ride for the user, reduced sales for the company. See examples below of the inconsis

The solution: Bring ecom under your primary Web property so users trust you - leading to increased online sales. Plan for ecommerce within your information architecture.

Example of a third-party payment system

Clicking on conEdison's "Pay by credit card" option takes you to a third-party company that handles the payment. The transition is jarring for several reasons:

• New URL doesn't mention conEdison
• NCO logo replaces conEdison's in the upper left
• New color scheme
  
• New privacy statement.
  

The payment agency does responsibly provide a link to their own privacy policy and disclaimer that they are a third party that handles payments. Yet, other than the inserted conEdison logo, it's difficult to tell that this page relates at all to conEd's parent site.

Ecommerce is alive, usability problems still rampant

Clearly ecommerce is far from dead - it leads growth in retail even during this global economic downtown. Yet usability problems, which negatively affect sales, are rampant. Why has this problem remained with us since ecommerce first began?

In August of 2001, the dot-crash was roaring, about to be amplified by the events of 9/11. Early obits for the Web were popular. So it was that usability eminence Jakob Nielson asked, "Did poor usability kill ecommerce?" At the time, Nielsen conducted usability testing of 496 users on 20 US-based sites; failure rates were 56%. Nielsen also estimated that 79% of the sites could increase sales by improving usability. What has changed? Well, ecommerce is no longer an experiment. From Internet Retailer: "Sales of the Top 500 online retailers grew 11.7% to $115.85 billion in 2008 from $103.69 billion in 2007 while Internet Retailer estimates the total retail sales market grew 1.4%."

So ecommerce usability problems have gone away, right? Wrong. In many of our usability tests (2000 - 2009), we've observed all kinds of users attempting to purchase online, and we've watched them abandon transactions for the same old reasons:

• Clunky or incomprehensible checkout process
• Brand inconsistency between the parent site and the ecommerce application
• No evident privacy policy
• Confusion returning to shopping after placing a purchase in the shopping cart
• Failure of shopping carts to retain user choices
• Asking users to complete extensive forms prior to checkout

These are user task path problems, with a dose of emotional intelligence in design sorely missing; look at ecommerce interfaces for a myriad of more straightforward (and unacceptable) UI problems that undermine users and impact sales.

There is no replacement for information architecture, user interface design, and user-centric task sequence design as processes required for usable ecommerce platforms. Companies that took the time to stabilize their ecom platform before adding a plethora of products are the winners.

If your business model relies on online sales, plan to increase those sales by conducting usability testing and investing in solid information design. Time to stop guessing.

Out-of-the-grid design - when is it a good idea?

My longtime partner in Internet adventure, @jackpowers, posted this today:

@Jack Powers Eff usability, I love great design. RT @onextrapixel Characteristics of Less Conventional Websites w/Examples http://bit.ly/19iDHq

So I checked it out. Jack's got excellent taste, and the examples provided by Charlotte at onextrapixel.com are gorgeous, modern, compelling.

My take: I adore great design. As a fine arts major, I remember grumbling about why we had to master still life before creating abstract works, about the excruciating color-chart exercises. Eventually I understood that I needed those fundamentals before I could truly abstract.

We'd love to see a Web full of sites like those at oneextrapixel. But as a working (not theoretical) information architect, I must say that the best candidates for this sort of visual design are professional visual designers and hip development shops (there are quite a few portfolios there) and, of course, entertainment, where the non-conventional approach has always been appropriate.

Information-rich Web sites need to have gotten Web 1.0 right before moving to the unconventional Web site. This means branding, strategy, user priorities, information architecturem user interface design. All the boring stuff that creates a truly usable Web site. Then we can (if justified by mission) move on to abstraction.

Jazz musicians must master the piece before they can improvise. We look forward to working with clients that have already mastered the Web fundamentals so we can take them to the next level. Until then, our job is to get them to master the fundamentals. And to keep an eye on beautiful design, so we can lead the there when the time is right.

How "nimble" should a Web site be?

Talking through Web strategy with a client today, we consider just how "nimble" a Web site should be - how frequently should a Web site redesign, based on changing mission priorities?

This is a particularly difficult question for large organizations with so much to say and no easy way to say it. Web sites that bend too far in the direction of the mission du jour may neglect content that is not part of that mission. The result for unsuspecting users may be "information paralysis," an overwhelming amount of content that could have been useful - had the user route been planned for those residing outside of the current mission priority.

Large associations and nonprofits frequently struggle with the challenge of meeting the needs of multiple constituencies. The solution is not to change the entry experience every time the organization takes on a new initiative; the solution is to create an information architecture that will continue to support known and ongoing constituencies, while creating a structural option that accommodates the changing initiatives of the future.

The time for "nimble" is in the planning and prototyping stages.

Agile development and UX: The heart of the matter

Re my recent post on developers and usability:

If the user experience practice in your company was weak before Agile, Agile development isn't going to help things.

If your user experience practice was strong before Agile, it'll remain strong after Agile, and evolve to adapt.

So says Jeff Patton, and I couldn't agree more. No matter how you approach user experience, you must set aside time and resources for it. It has to be in the company DNA - even if that value must be ported from an offline discipline.

Agile development provides fertile ground for user experience design. It may not provide as much support for usability. Regardless, user experience design - most notably including user-facing information architecture and usability testing - must be accounted for in budget and schedule.

Do bad things happen when developers design user interface?

We work with some incredibly talented development teams around Web sites, Web applications, and software projects. Unfortunately, teams missing a user experience specialist often find that they cannot come to consensus on what the user interface should look like, or what it should do. This leads to a sub-optimal user experience, with the common result being that powerful software functions are invisible to the user. That wastes money - and obscures the excellent work of your team.

The cost to business of skipping the critical planning stages - which include user profiles, usability testing, information architecture, user interface design and task sequence design - is incalculable. Skipping these steps means endless hours of expensive meetings where everyone has an opinion, but few have real data.

Stop for a moment and consider the areas of expertise your employees own today. Is there someone on staff with formal training in user interface design? Information architecture? Task sequence design? If the answer is "no" or "sort of," it's time to start adding up the cost of all the hours spent in meetings with no consensus, so you can reallocate those funds to usability training or professional services that will bridge the gap.