GarnishBar is a project I work on with fellow friends. Succinctly, GarnishBar helps you find and make great cocktails.

Situation

Lots of sites have recipe databases with hundreds of recipes. But the quality can be hit-or-miss. We list fewer recipes than our peers. Our focus is on quality and “discoverability”.

The site

GarnishBar is a deceptively simple on the front-end. Behind the scenes, a highly custom CMS maintains complex taxonomies for drinks, ingredients, glassware and other recipe data. Nearly everything is deeply related, which powers a faceted navigation system. Users can discover drinks based on ingredients, keywords, certifications, and more. A mobile version allows users to search from the bar. Unbeknownst to the user, the site delivers different ad configurations dynamically, tracking the effectiveness of format, size and placement.

At the time of this writing, GarnishBar is in its first release. We’re working hard to build in user accounts, recipe submissions, recipe cloning, tasting notes, ratings and other social features. Our hope is that we give people who care about mixology the inspiration and information to make great drinks.

My involvement

Perhaps different than most of the projects I’ve worked on, this one has the blurriest lines. Officially, I’m the UX guy. However, every decision whether information architecture, technical, design or interaction-related was vetted among the team members.

Situation

The American Institute of Certified Public Accountants (AICPA) wanted to connect with its early pipeline audiences–high school students. Their existing site wasn’t cutting the muster, so we completed a complete redesign.

Involvement

For this project, I conducted formative user research with students in our audience. In a series of one-on-one interviews, I gained insight into their perceptions, needs and challenges. Later, I conducted summative research to identify which of our concepts resonated best with them. With a team of designers, developers and copywriters, I developed the Web strategy. Later in the project, I built both paper and dynamic prototypes  to define the site architecture and  interactions.

The Web site is designed to give students a taste of the profession. Rather than encumbering them with all of the details, it focuses on giving them just enough information to think prospectively. It answers the questions “Why would I want to go into accounting?” and “What can I do in this profession?”

We found that many stereotypes still exist. Students who have not been exposed to accounting tend to view its practitioners as bookish, isolated number crunchers. Nothing could be further from the truth. Throughout the site, we break down these misconceptions. This could be by profiling actual CPAs in interesting industries–or letting them know that even a Butthole Surfer is a CPA.

Involvement Highlights

• Web strategy and feature definition
• Technical specifications (Facebook Connect, Telligent Community Server)
• Information architecture
• Prototyping
• Interaction design
• User testing strategy and execution: concept and preference testing
• Visual design
• Content strategy
• Quality assurance testing

Results

The site receives hundreds of visits per day, with time-on-site averaging over five minutes. In two months, with no promotion, we saw nearly 4,000 new accounts created. We’re constantly examining the analytics and updating popular content and features to improve performance.

Situation

UNC-Chapel Hill’s primary Web site had been unchanged for nearly a decade. There had been some incremental improvements, but these repairs did little to improve usability or branding. We completely redesigned this Web site, with a particular focus on telling the University’s story.

My involvement

• Web strategy and feature definition
• Information architecture development and testing
• Information graphics
• Accessibility strategy and tactics
• User and stakeholder interviews

Situation

The American Institute of Certified Public Accountants (AICPA) found that its Web site did not meet member needs. In addition, its architecture created operational inefficiencies and prevented AICPA from staying current with technology.

My involvement

Working with stakeholders throughout AICPA, I did the following:

• Web strategy: Defined the operational goals for the site and the features to support it.
• Information architecture design and testing: Conducted a full content inventory and reorganized the content and labeling based on users’ mental models–not on organizational structure.
• Wireframing and prototyping: Wireframed all page templates. These are designed to scale easily to different types of content and work well with a Sharepoint / Endeca powered environment. In addition, they provide affordances for the social networking features that are planned for future releases.
• Usability testing: Conducted remote card sorts and scenario-based tests to refine and validate the architecture.
• Content strategy: Wrote the strategy that identifies the objectives for the new Web site, the rationale for architectural decisions and tips for writing and organizing content for the Web. Produced training materials for content creators.
• Design concepts: Created an early round look-and-feel concept.
• Various information graphics: Designed graphics to help clarify complex architectural and procedural concepts to non-technical stakeholders.

True Tale

We encountered a few emotional debates about architecture and nomenclature. Some stakeholders had trouble accepting the Web site should not necessarily reflect the org chart. In these cases, data from card sorting and usability testing gave us independent evidence to base our decisions. This limited arguments based on opinion and helped us make decisions on behalf of the user.

The Web site is scheduled for initial release in early 2010.

Situation

The James B. Hunt Jr. Institute for Education Policy and Leadership is a non-partisan organization that connects policymakers with education research. They host a series of conferences for state governors.

In our user research, we found that we could distinguish the site by meeting the needs of legislative staffers. These individuals are called upon by governors and lieutenant governors to assemble education performance data. This group was not served by the old site.

Features

The site offers a number of valuable features. These include:

• A consolidated database of education performance data. Staffers can immediately see how states compare with one another and find the original data source.
• A relevancy algorithm to serve sidebar content related to what users are currently reading. If they are reading an article on school scorecards, they will see related events.
• A rich metadata model that allows the institute to provide filtered views of their site content based on audience, topic and state. For example, they can create a dynamic page with content just for North Carolina teachers interested in standards of learning.

My involvement

• Perform competitive, business and gap analysis
• Conduct user interviews and develop personas
• Define the Web strategy and site features
• Develop the navigation and metadata information architectures
• Produce wireframes
• Conduct quality assurance testing

Situation

Recruiting within management consulting is fierce–especially among the Big 4 consulting firms. Deloitte knew that to attract and retain the best talent, they needed a different approach–one that was more personal and human. After conducting user research and ideation, we produced a site that focused on the unspoken questions that recruits don’t ask–but want to.

My involvement

• Conducted interviews and developed personas.
• Developed the Web strategy and supporting features.
• Created sketches, interaction diagrams, information architecture and final wireframes.

Additional

Recently, Deloitte consolidated all recruiting sites under a single look-and-feel. The best of our features, however, were retained.

Scenario

Engaging hundreds of users in a collaborative brainstorm is challenging. People have different schedules and live in different timezones. And some are reticent to offer their opinions, for fear of sounding dumb. Further, it can be hard to see the overall pulse of a community.

Features

Ziggurat is a Web-based application that allows users to engage in a brainstorm on their own time. It’s part survey tool, part forum, part idea miner. Moderators create questions for users to discuss. Users provide their answers and can rate / comment on others’.

To help channel the discussion, moderators have access to an analytics engine. This provides insight into user engagement and trending topics. Based on this data, they can focus the discussion around the ideas generating the most interest.

My involvement

• Defined the features
• Created an interactive prototype for the offshore development team
• Designed the user interface and look-and-feel

About the name

Ziggurats are stepped pyramids common to Mesopotamia. In some cases, their “steps” are nothing more than a wrap-around inclined plane. We chose this codename because it alludes to the iterations and progressive focus that occur during “Ziggs.”

Problem

Usability testing for Web sites is usually a one-off event. Every now and then, tests are conducted with great fanfare. While this is good, it misses the opportunity to constantly measure the usability of an application. Additionally, this approach makes it difficult to see longitudinal effects–such as how site users learn over time.

At the time, competing products such as Keyhole were grossly expensive and challenging to implement. What I wanted was “Google Analytics for usability tests.” Autopilot was my answer to this business need.

My involvement

• Assessed competitors
• Estimated development effort and conducted an NPV analysis
• Developed a interactive prototype
• Defined the business and technical requirements
• Created the visual mockup of the interface
• Collaborated with a producer and senior management on pricing models
• Performed a cost-benefit analysis

Where is it today?

The company I worked for was willing to bankroll development, simply  to have a differentiating service. This was good.

In the early stages, however, two competing services emerged that had attractive price points. After projecting future cash flows, I found that the break even point was too far in the future. I made the somewhat hard decision to stop work. But sunk costs are no reason to continue development.