Surreal Notions

Intuit initially hired me as a staff UI Designer to support TurboTax® Estimated Taxes.  Although I designed interaction for numerous projects while at Intuit, this project represents the majority of my time and effort.  I worked closely with a talented team of engineering, QA, support, domain experts, and product management personnel.

We conducted several usability and marketing studies from late 2006 to early 2007.  Each study focussed on different aspects of the user experience ranging from SEO to help the user find the application to specific tasks a returning user wants to accomplish.  Among the many learnings from this research, we found the majority of our users were elderly retirees.  In fact, the initial offering had substantial negative feedback around navigation and confidence.  Basically, users had trouble finding where the tasks lived and weren’t sure when they were done with the task.  We also saw high rates of abandonment during setup.

To mitigate these concerns, I started by redesigning the setup user experience from a 14 step process to 3 steps.  This reduced the abandonment rate and initial confusion.  It also allowed users to get into the product and use it before deciding if the product solves for their specific needs.

Once in the product, the home page redesign used a checklist metaphor to help guide the users to the tasks necessary for their specific needs.  The checklist or action list guided users to next steps required by the IRS, errors that must be corrected, and points of concern within their account.  Several variants and metaphors were tested during usability, but the checklist metaphor clearly rang with the majority of subjects.

I went on to simplify the filing process to a single screen from a 4 step process; simplified the account management and reports sections; and improved navigation and confidence within the estimated taxes calculator.

Although these improvements were met with resounding positive feedback, technical changes with a third-party component has forced Intuit to temporarily deactivate the product.  Plans are moving forward to work around the third-party component and open TurboTax Estimated Taxes up to the public again or imbed the functionality within TurboTax desktop products.

Designed interaction for two of four teams assigned to TurboTax® Business 2007.  While conducting several usability studies, I helped the team of tax experts improve form selection, expenses, and other areas particularly confusing or troublesome for users.  One thing to note here, the TurboTax Business product is built upon Intuit’s professional product.  It is a fairly complete solution, but still contains quite a bit of content originally written for accountants.

The form selection process was reduced to a single screen.  With the help of tax experts and a copyrighter, we improved the layout and reduced confusion between the different options.  For instance, a few rare instances require unusual filing practices.  Rather than making these possibilities differentiated selections (giving them equal weight to the more common options), we improved the help content and limited the user to simpler options.  Although this might create confusion for the odd-ball filer, our studies found the percentage of filers with those conditions to be well below 5% and a majority of them understood their situation and didn’t require help.  Usability indicated users not only found the correct form faster, but were much more confident in their selection and the rest of the product.

The single biggest problem small business owners have when filing taxes, assuming their books are accurate, is correctly assigning business expenses.  To improve this process, we decided to use a novel way of customizing examples in each expense section to better represent the type of expenses the user is likely to have.  For instance, if the user indicates his/her business is a manufacturing plant, expenses commonly associated with manufacturing are displayed in the respective expense sections.  It’s not likely a consultant has farming expenses, so we leave them off-screen in the help section and concentrate on travel and office expenses.  Usability firmly supported our hypothesis that users are more confident of choices when one or two examples exactly match their needs.

I look forward to picking up a copy of TurboTax Business 2007 when it becomes available to see how everything works for my business.

Great news everybody!  Philip Haine (Obvious Design, Steal this Idea) has invited me to work with some of his clients, because he’s taking a sabbatical to finish his book on product vision.  Philip and I worked briefly together on a couple projects at Intuit, and I’m really excited to have the opportunity to learn some more from him.  Philip has been a user interface designer for years and consulting for about 11 years.  Needless to say, I have much to learn from the Jedi Master.

Philip was the primary designer behind TurboTax Expense Pro (I had a bit of input simply because I sit 30 feet from the developers).  He also worked on MyCorp and TurboTax Business.  In fact, Expense Pro and MyCorp are the two products with the most potential at Intuit in my opinion.  Philip designed very solid interfaces for both applications.  Philip walked me through the terrifying complexity of MyCorp with ease… which is quite a statement considering the hundreds upon hundreds of offering combinations possible–each with legal and financial implications.

I should have my official LLC documents within a couple weeks, so it is looking up for the consulting gig.  With Philip’s generous help and a little luck, I’ll be able to get involved with a couple Silicon Valley projects before the end of the year.

I will be starting a short-term contract with Northern Natural Gas next week to help clean up an internal application and lay down some UI groundwork for a longer term consolidation effort.  Effectively a survivor of the Enron debacle (or maybe they were just purchased by one of the fragmented companies that came out of Enron), NNG supplies natural gas from Texas to Lake Superior.

I’ll be working at their headquarters in Omaha, which should be interesting.  I’ve heard good things about the work environment, but suspect the internal IT projects are underfunded with poorly implemented incentives around the business unit… which is pretty status quo for IT in my experience.  Either way, this will keep me busy through at least the first week or two of September, which is good.  With a little luck I’ll have the consulting business up to speed by the time this contract runs out.

I am in the process of forming an LLC to start independent consulting.  This, with a little luck, will help me put my skills to the test.  Most importantly, I hope to get involved with some interesting projects around the country that push interaction design to the limits and test my ability to manage product vision.

Sorry about lack of content on this post, but I’m excited to see where this leads.

Designed interaction for an ACI Worldwide management dashboard.  I worked with a small team on a prototype mockup, which we then presented to upper management.

Confidentiality prohibits me from going into detail; however, I did come up with a novel graphical interface, which could dissect data across multiple dimensions.  The system also integrated group-work and sharing features to help communicate analysis between individuals and management teams within the organization.  There were two competing presentations; however, management was genuinely impressed with our solution.  Unfortunately, all of the solutions presented would require serious infrastructure upgrades before practical to implement.  Hopefully they will get there some day, because I think any of the solutions presented would improve efficiency and efficacy across ACI… or any Fortune 500 company for that matter.

After spending a bit of time in Special Projects, we found a majority of our projects were database related and could be optimized with an improved component architecture.  My partner, Aaron Burke, and I set out to design and build this portal system with a central database to mitigate data accuracy problems.  I collected requirements via user interviews, designed the product interaction, and developed the database structure and procedures to support the applications.

We began by interviewing users in various units and positions around base.  We found, much to our dismay, that most units had rampant data accuracy problems.  One unit alone had over 250 rogue databases, mostly Access databases written by non-technical personnel.  Keep in mind, this particular unit only has about 700 personnel total, so every three people effectively had their own database.  To make matters worst, the inherent bureaucracy of the military convolutes who owns the information.

Our first complete application was My3215 Online.  This basic workflow and project management application took advantage of our central repository of information, which allowed anybody on base to initialize a request with minimal, if any, setup.  The request was then automatically routed to the appropriate authorizing member, who also applied a priority (Critical, High, or Nominal).  From there project managers could direct the request to any shop that needed to do something with the request.  Annotations, logs, and attachments were all became part of the request.  The requesting user could check on progress whenever they wanted without distracting project managers with phone calls.  By limiting the number of critical and high priority requests available to any given authorizing member, we helped reduce the number of ‘emergency’ requests made.  The system also sent email updates to users initiating the request and any agent lapse in required actions, which helped reduced the number of requests left incomplete due to a missed step.

The flexibility of the system allowed project and shop managers to easily stay on task and maximize workforce efficacy.  Although accounting methods made success difficult to calculate in terms of dollars, the communication group commander estimated $150k-$300k were saved per year.

I wrote the following documents as training/help material for Base Portal User and My3215 User.  The military generously gave me permission to post this information because the application was decommissioned last year.  Apparently, they no longer had resources to support the application.

A team of five low-ranking airmen were assigned the daunting task of finding a replacement for the 40+ year old Strategic Automated Command and Control System (SACCS).  Loving a challenge, I couldn’t resist joining the team.  The current system’s maintenance costs are spiraling out of control as parts become non-existent and service becomes more difficult.  We quickly decided the best route would be a modern approach to computing that takes advantage of common, inexpensive, disposable parts.

We began by interviewing a few key personnel from the SACCS development and maintenance teams.  There are very few people that still know how to work on systems this old, so our choices were limited.  We also dug into the SACCS documentation, but quickly found that 30 years of spaghetti code and shifting management priorities made the documentation convoluted and, frequently, contradictory.  To make things more interesting, the NSA security definitions had substantially changed in recent years.  After a few weeks of research and heated discussions, we had a good idea of what the system needed to meet user and security requirements.

We proceeded by researching the handful of vendors in the world that might be able to provide a baseline secure system, which uses modern technology (i.e., standard Intel based server/client architecture).  We invited the two most likely vendors to visit our operation and discuss their thoughts on the challenges and options available.

Ultimately, I compiled all of the information the team had gathered into an executive summary.  We proposed a phased plan of implementation that would cost $3M – $7M and take 2-4 years.  The wide range accounted for possible development challenges related to integrating communication across older segments of the network.  Of course, the cost is almost trivial compared to the $400M – $650M in savings created by dramatically reducing maintenance, overhead, training costs, and footprint.  I was informed the Chiefs of Staff reviewed the plan in late 2005.

This database was originally an Access 97 database written by a member of the medical staff at Offutt’s hospital.  It was used internally by a dozen or so people at the hospital.  Then upper management saw it and decided they wanted to leverage it against the entire base… and so I was given the project with an initial release planned mere days later.

The user base was going to increase by about 100 fold.  Not to mention additional functionality required and a generally poor baseline.  To top everything off, I needed to make it HIPAA compliant.  Needless to say, this was an interesting challenge.

After a mad dash to break the application into two components and rewrite each component, I managed to get a working copy up and running before the base-wide readiness inspection.  This application was specifically noted by both management and examination teams as key during the inspection.  Eventually, I designed an improved version that used a full DBMS and web interface; developed more effective reports for differing levels of management and need; and increased security and compliance.