Recently, I made several big announcements on this blog. One was that the company is now called Thrill & Create LLC, and another is that the company’s new website is now online at AmusementUX.com.
The previous site
The initial work for AmusementUX.com, internally abbreviated AUX, started almost two years ago. The company’s old website, at DalandanConcepts.com (pronounced “dah-LAHN-DAHN”), was the first new website I had designed or developed in about 9 years. (I had built several music-related websites, which are mercifully no longer online, over the 5 years before that – using a very old Mac version of Dreamweaver.) The Dalandan Concepts site was entirely hand-coded HTML and CSS using the now-deprecated 960 grid system. In other words, it was intended to only look good on desktop computers: not a good selling point since I was trying to sell usability consulting services using that site. Last year, I would take part of a weekend to make the site responsive so that it would be at least somewhat usable on mobile devices.
Preparing for Project: Amusement UX
I started planning the site’s replacement almost as soon as I launched it. While researching hundreds of amusement-industry professionals on LinkedIn, I generated a persona spreadsheet. This was relatively preliminary user research compared to what I do now, but this spreadsheet informed the design of the rest of the project.
I initially knew the site as “Dalandan V2” (version 2). I began by developing a desktop-first wireframe and later replaced that with a mobile-first wireframe. Fortunately, I was then busy with client work for a while. Many incoming phone calls during that year, in which people mispronounced “Dalandan” showed me that Dalandan was not a good word to use in the company’s name, whether I had nice pictures of dalandan (a type of fruit which I ate in Southeast Asia) to use for the website or not. Ultimately, what the site needed to sell was design services, not food.
So aside from making the existing site work on mobile devices, buying AmusementUX.com, and identifying a WordPress theme, I did not do any additional work on replacing the company’s site in 2013. I decided to shelve the project until mid-January 2014. Another big decision was to use a daily Scrum process (adapted for a Scrum team of one) in order to design and develop the new site.
Amusement UX: design and development in full swing
The project’s overall structure, including both design and development, consisted of some preparation work, 7 iterations, and 4 spikes.
This was a brief iteration before work on the Thrill & Create site truly started. I installed a Coming Soon page and wrote some brief copy describing our business. Design work during this phase was fairly minimal.
During this iteration, I performed my first round of ideation. Hundreds of ideas were then pared down to a much more manageable set that fit the site’s primary personas, and I did sketches and wireframes. My tool throughout the wireframe/prototype process was Axure RP 7. I also wrote preliminary copy for several pages. On a client project, I would have wanted to do a round of testing with users at this point.
At this stage, the homepage was much longer than it is now. Services, blog posts, and the About section were described on the homepage. An initial Process page, not yet reviewed with fellow designers, was already part of the site.
Iteration 2 fleshed out more of the ideas for the site’s Services page, Process page, and the homepage. The homepage, at this stage, really aimed to establish the site as an authority regarding usability and user experience. It also did more to sell users on responsive design.
By the end of iteration 2, I had made medium-fidelity prototypes of most of the pages in the site. I had also run through Zurb’s Design Triggers list and incorporated many of those ideas. This iteration ended with a round of short usability tests to gauge users’ first impressions of the site.
The first impression tests told me to revisit the layouts of the homepage and its hero area. I generated 4 hero area ideas and 8 homepage layout ideas and created wireframes and medium-fidelity prototypes of each one.
I then ran a survey wherein I paid many users to tell me what homepage layouts provided the strongest, most professional first impression. I chose pairs of ideas to compete against each other in this. Two out of four pairs did not have a clear winner, so I created layout ideas 9 and 10 as hybrids/replacements of these 4 other ideas.
Iteration 4 incorporated feedback on the hero area surveys and a next round of homepage layout surveys. These resulted in some modifications to the hero area and the homepage. At this point, I ran a set of longer usability tests on the prototype and triaged their feedback. During this stage, I was also working on some business strategy options related to my usability evaluation service offerings.
The longer usability tests gave me a wealth of valuable feedback. Among other changes, I continued to sketch new ideas for the homepage layout and created new variants of the hero areas. I also created new ideas for the Process and Services page, wrote their copy, and created two more comprehensive prototypes of the Why User Experience Design Instead of Web Design? article. During this stage, I was also making preliminary choices for the site’s typography.
Several users in this round of usability testing did not like the color palette which the site was using at that time. Since changing the entire color scheme of a site involves widespread changes and it is not (as of this writing) a simple process in Axure RP, I created a separate sprint spike to work on this. By then, the site was already well into its development process. The Axure RP prototype with the new color scheme gave me a reference for how the site was supposed to look after my code changes.
Iteration 6 began in early May with adding content to the site, which was still hidden by the Coming Soon template. I started writing a custom CSS file, which eventually grew to well over 4000 lines of code. I spent most of this iteration working on the custom CSS and its associated work items. There was also a support issue with a vendor which took weeks to resolve. It pushed back the launch of the site due to the issue’s severity and the amount of development effort I had to expend to formulate an acceptable solution. Business owners wear many hats, indeed.
Spike 6.5, which didn’t meet a standard Scrum definition of a spike as well as I wanted it to, was mainly used for fixing bugs with the site which I had found on mobile devices and for starting trials of the fonts I was going to purchase, in advance of testing the site again with users.
In Spike 6.5, I created Axure prototypes of the homepage’s layout at more widths to show what a working version of the site would look like at those widths as I worked through the bug list.
At the end of iteration 6.5, I conducted another round of long usability tests, including several tests with other UX designers. This, again, resulted in much valuable feedback.
At this stage I implemented an element collage to ensure a more consistent look and feel across the site. Element collages are now an artifact I produce much earlier in my design process.
In development, iteration 7 also involved making sweeping CSS changes because I changed the site’s font pairing in response to user feedback.
Spike 7.5 included more mobile bug fixes and many deployment tasks. This spike ended on August 21, 2014, when I soft-opened the site that you see today with a placeholder company name. After more branding surveys, I began the process to officially rename Dalandan Concepts to Thrill & Create.
Next on the agenda
Here on the Thrill & Create site, you may find information about who we are and what projects we are doing. Our Facebook and Twitter pages will have further, more frequent updates from the intersecting worlds of user experience design and amusement.