connecting*the*dots

(originally uploaded by wmchu)

I had a meeting yesterday over lunch with a client of mine — Louis Bowles of locally based Louis’ Healthy Breads. Louis, John Ford and I are in the midst of planning a web strategy for his small business; an execution that will include the implementation of a blog and a redesign of his e-commerce enabled site.

At first glance, the scope of the project seemed extremely manageable and somewhat simplistic — essentially, it’s a blog with a product template and a tie-in to PayPal’s fulfillment processing. Over the past few months, though, I’ve come to recognize that the challenge of the project isn’t in its technical complexity or feature set, rather, it’s in pulling off the personable nature of the brand.

When Shiny, Shiny Design Doesn’t Work

Louis didn’t start his company to make short-term killer profits (though, I’m sure he wouldn’t turn them away); Louis took his creation to market because, quite honestly, it saved his life and he feels that people need great tasting, healthy aternatives in their diet.

My challenge is translating that degree of authenticity into the experience design of his web site.

Chris Fahey just posted the sixth installment of a brilliantly focused series entitled Class and Web Design. The ensuing conversation regarding the impact of class on the output of design is fascinating (read the posts and comment threads; I could never do it justice here). While I doubt “class” is the proper signifier for categorizing Louis’ concerns about coming off like a cold, mass-produced food industry giant, to Chris’ point, a company’s outwardly focused brand position is intrinsically tied to their target (and if they’re responsive, their actual) market.

Just take a look at the redesign of (and the conversation surrounding) the New York Post for an example:

As The Post shoots for local readers with specific desires of sports, entertainment and gossip coverage (all big type desires), Louis’ brand needs to register as if he’s your next-door neighbor — someone who deeply cares about your health and just so happens to have a busy kitchen at work to help you make sensible, delectable choices.

Does that mean the design of the site needs to be overly pedestrian? Not at all, but as with the example of The Post redesign, it needs to remain true to its core principles of Louis being Louis. The visual language of the site needs to reflect his engaging personality, while presenting a strong enough degree of credibility for people to latch onto.

The beauty of this project is that there’s enough room to play and enough leeway to iterate.

Design Basics

Subtraction is the standard approach to good design and focused communication — particularly within the interactive medium — so it’s essential that we reduce the brand to a distinct visual language and behavioral model that clearly communicates with the people who benefit the most from the product.

And because we’re adding the communicative element of the blog to the mix — somethat that will improve our responsiveness in meeting customer needs and steering the direction of the brand — the last thing we want to do is bury the product behind a conversation about the product.

With a brand as personal as LHB, we’re going to have to tow the line between authenticity and credibility. Prioritized simplicity, both visual and organizational, is key in making a step in that direction.

FORTUNE COOKIE: Sometimes the biggest challenges come wrapped in the smallest packages.

If you’re an interaction designer, think about the process of generating design personae while listening to Gladwell.

Back in February, NBC made a completely bonehead business move by making YouTube take down the hugely popular video short Lazy Sunday. My instant response was to fire off a salvo at NBC for being old media ogres (NBC: We Get Web 2.0… Sike!) and not working within the limitless parameters of the web to strike a business deal that suits their needs to protect their copyright, while allowing us to continue to enjoy their content when we want and how we want.

Well, today NBC announced that it’s embracing a few of the ideas I previously lobbed into play:

[…]

“Under the deal, YouTube will create a separate channel for NBC video, so that visitors can easily pull up the half-dozen or more items that NBC plans to offer at any given time. It will be similar to channels that other companies, filmmakers and everyday users create.

[…]

NBC and YouTube officials acknowledged the possibility that fans will reject the clips if they appear simply as promotions, but YouTube co-founder Chad Hurley said fans would likely embrace the video if it is compelling and not available anywhere else.”

[…]

Promotional video is somewhat of a start — I suppose you can’t expect major change from a major television network without them testing the water first. Give the experiment a few months; if uptake begins across numerous types of unbundled content, I’m sure they’ll be banging on YouTube’s door, attempting more creative ways to “let” people upload their content.

Affecting The Interface

In terms of the user experience, I only ask one thing of YouTube: please refrain from creating a pulldown of “channels” on your interface.

Asking people to assign ripped video to a “media channel” in the upload process makes sense:

• It alerts you (YouTube) to content that needs to be assigned a “shared monetization flag” and
• It automatically assigns network metadata to the video object to help people finding content they desire

Balancing the two-way participation of a user base with the business opportunities of old media is a difficult conversation to manage and execute, for if you transform your main interface too far towards the navigation of paid-for, primary channels, the entire participatory, community vibe will begin to deteriorate.

Remember, your brand is YouTube.

Peter Morville, Information Architect.

Morville classic quote: “Information Architecture: A balance of art and science.” Risk taking, creativity, listening, trial and error. Designers, writers, developers, etc. are all practicing information architecture techniques (i.e. Microformats)

Different types of domain and users need different types of information architectures.

Search is a System

• User query ->
• Search Interface (Query language, builders) ->
• Search engine ->
• Content (metadata, CV) ->
• Results (Ranking and Clustering Algorithms, Interface Design

Searching is not only finding, but learning (discovery)

Findability
Can people find your web site, find content in your web site and find content despite your web site.

Shifting Gears: “One foot in the past and one foot in the future” What are the longer term trends?

“A wealth of information creates a poverty of attention” - Herbert Simon

You know what? Peter is too eloquent for this live-blogging crap. Go buy the book; it truly is a great read.

Adam Greenfield is dealing with Godly AI interfaces.

What is ubiquitous computing?

Well, what happens when computers get cheaper, faster, better? They become invisible, but all around us. The possibilities to crunch concepts, data, information explode. We move into a post-graphical user interface, from gesture to voice.

Multiple users, multiple spaces. Moving away from the one-to-one paradigm.

Human behavior and ubicomp become as one. “The activation process dissolves away into the behavior of people.” Ubicomp is already social. Once devices become ambient, social interactions can meld into a contextualization of backgrounds, ideas and relevance.

• It’s present at the level of the body to world interface, as with data captures of physical movement to track, say, the range of motion of an elbow.
• It’s present at the level of a particular space — a room — to a processor reading the reactions of the room.
• It’s present at the street level, reacting to movement on the street and surveillance of social interactions

Ubicomp can crunch a variety of input or data, all passed through a relational database to construct information or entertainment for digestion.

People live life in real-time, while ubicomp works with their behavior to support their needs/desires. Space is never neutral, as the politics of position can be taken in numerous degrees.

Ubicomp is now. Why?

(His cell phone goes off ;)

The digital home is the next big market and the future is structurally latent. (Crazy meta-meta-meta tagging in the real). Also, public safety comes into play. Post-9/11 mentality has crept in with, “Reduce the publics fear, reduce access and monitor activity.” We need to engage in ubicomp to control our destiny and the degree of misery which could be on the horizon.

Locus of attention disappears with ubicomp, so troubleshooting the invisible become a cognitive challenge. Signage is incredibly important to navigate the explicit behavioral captures of our implicit progression through our day-to-day.

“The challenge of implicitness is… an ethical challenge.”

5 guidelines of designing for ubiquitous computing

1. Ubiquitous systems must default to a mode that ensures their users safety (physical, psychic and financial). Graceful degredation moved towards a default to harmlessness, based on cultural definitions.
2. Be self-disclosing; ubicomp must contain provisions for immediate and transparent querying of their ownership, use, capabilities, etc. Seamless interaction in physical spaces must be optional, as ubicomp could invade the privacy of individuals. “Seemfullness with beautiful seems.”
3. Be conservative of face; allowing people to save face. Ubicomp must not unnecessarily embarrass, humiliate or shame their users. Humane interfaces must be taken into consideration, especially while designing the experience of invisible ubicomp systems.
4. Be conservative of time
5. Be deniable; allow for the opt out of the program at any time. Alternatives should be provided to people who want to avoid these systems.

Disclaimer: This is live blogging; all quotes are paraphrases.

As social tagging begins to catch on beyond the early adopters, content and commerce domains are opening up their information architectures to empower their consumers to tag, creating exponentially greater degrees of faceted, semantic relationships between their information objects.

Amazon is already in the lead to extend this open paradigm into the commerce space with object tagging and Mechanical Turk (a program which could seriously disrupt peasant-class wage pay around the world). Amazon’s past innovation isn’t a guarantee for future success, but their recent moves prove to be a good sign.

How Social Tagging Works

Folksonomies change the dynamics of generating useful index pages by centralizing human perspectives expressed through single or compound descriptive terms into navigable indexes. It’s the equivalent of a dynamic, open-ended thesaurus, eliminating the need to manage the static creation of valued relationships, as co-occurance stitches together threads of information like newly created and evolving synapses in the brain.

The usefulness of these visible, semantic relationships to the person searching for specific content or products is quite possibly the most sticky form of extended discovery not generated through database algorithms.

I mean, forget dropping out of my mental model to browse topical navigation or stopping to search for an explicit term or phrase; when I engage with a domain such as flickr or del.icio.us, my desire to stay within the domain is increased simply because the language I use to define my world through tagging simultaneously allows me to peer into the world of like-minded folk (ergo: folksonomies).

Tagging creates community through the overlap of perspective.

While this extends conversation, it can also impact the sales potential of commerce sites by adding another layer to collaborative filtering, which Amazon has already acknowledged through their advancements in tagging. Now, extend this concept further into the realm of consumer contributions with industry and one can envision the incentive for business to slightly open their gated approach of mass manufacturing in this age of personalization, allowing customers to participate in defining what a company produces by simply tagging their existing objects.

• Tagging builds community
• Tagging increases the findability
• Tagging can give customers a transparent stake in the process of creating services/products/content

Back To The Interface

Try thinking about tagging interfaces on a few distinct levels:

• Interfaces which display common tags from across a particular domain need to be designed to maximize their semantic relationships.
• Object-level interfaces need to be re-crafted to both accommodate the display of previously applied personal tags and tags applied by the community.
• Management screens, which can give ownership of personally applied tags to the people that spend their time generating them, need to be compiled from contributing domains across the web for individuals to manage and, potentially, collect residual dividends related to sales generated from exposed tags.

I recently stumbled across an interesting site that leverages the API of del.icio.us tags. Kevan Davis created extisp.icio.us to scrape user tags and visually represent them using only words or images:

My good friend, DeWitt Clinton, created Delancy, which leverages the open nature of del.icio.us, providing an enhancement with the ability to manage tagged objects by personal click-through popularity:

Kevan’s enhancement focuses on re-presenting information in a way that presents our constantly evolving association with the world outside, while DeWitt’s enhancement focuses on adding feature value, assisting us to quickly find our most used bookmarks.

This type of innovative, open source development reflects the same type of creative energy that non-developers posses — people that are becoming hooked on tagging, hooked on participation.

Sharing Interfaces, Creating A Usable Web 2.0

Now that Silicon Valley is reaping the rewards of innovative open source development—observing hundreds of prototypes across numerous types of applications—how long will it be until these companies begin to act in a similar fashion? Yes, I’m talking about open collaboration.

TypePad enables me to tag my posts by assigning categories, but the management screen is a simple list, one that doesn’t allow me to easily create more manageable sub-categories (I’d probably group my tags by proper names, places, titles, descriptors, etc.). Mena, it’s becoming painful for me to manage my 200+ tags; how about TypePad teaming up with del.icio.us to use their management screen?

del.icio.us does many thing well, including their flexible interface for managing tags by give user created groups of tags nicknames. So simple, but so powerful. Why aren’t domains like TypePad, flickr, Flock, etc. bartering with del.icio.us to leverage this successful interface—one that thousands of early adopters are already using and loving — while providing their own best practice proprietary interfaces or code in return?

This level of collaboration amongst businesses is an example of what would allow companies to focus on developing more focused innovation, enhancing development cycles, reducing resource allocation and most importantly, providing best practice consistency across applications where possible. Toyota recently leased the technology of its Hybrid engines to Ford and other automakers.

How much quicker would a usable and useful Web 2.0 network be created if companies operated in such a manner?

The collective intelligence of humanity seems to be amped to contribute. Are we ready for them?

For the past three weeks I’ve been using Measure Map pretty religiously, trying to get a feel for its depth to see if it’ll be useful as a tool for me to use moving forward. My first review touched upon the usefulness of the features, but admittedly, it was much more of a review of the presentation. After pounding on it some more, I’ve a few more thoughts on the service (remember, this is still an alpha release):

Searching For Search
Measure Map presents the search terms that led visitors to my site from three major search engines: Google, Yahoo! and MSN. Below are the terms the were used between 11/05 and 11/20:

68
boondocks
bo peabody
bush crony appointments
bush lies
"contextual column"
"courtney bolton" new york
"David Reid" Baghdad
Dick Chaney and FBI Leak
DOTs
download ofoto
DUMB AND DUMBER
efrat yardeni
evangelic green card
"farrakhan"
farrakhan syracuse university
"Free Flow of Information Act"
Greensboro Troublemaker
Hadj guestbook 2005   
haiku George Bush
impeach Bush Now
javol
"jon stewart" + crossfire
Louis Farrakhan Rosa parks funeral
measure map
"organ failure and death" bush torture "new york times"
"navy seal"
newsbusters
revolution america
"Rosa Parks childhood"
"solo journalism"
sony apology 2005 for compromising PC
Visual map of shield law placement

While this feature is a common stat in an analytic tool, the data display isn’t complete:

• Technorati, Icerocket, A9, etc. query hits are listed elsewhere as links
• Image query hits don’t even show up

MM has a cleanly designed interface for displaying terms which originate from specific search engines, but it doesn’t include terms that originate across all search engines. If Technorati, Icerocket, A9, etc. can be presented on the link page, they can just as easily be presented on the Search Engine and Search Terms pages.

MM also differentiates text queries from image queries for no apparent reason. When I pause to see where visitors are coming from and formulate my understanding of why people are coming to my blog, images queries round out the story. Unless there are technical reasons for not presenting all search terms in one section, this should be a no-brainer enhancement.

My Blog Is More Than Just Posts
The number presented in the Posts icon on the homepage doesn’t equate with total page views (a common data point across all analytic services). I understand that Adaptive is trying to keep this simple—reducing page views to post views is one way to do it—but I’m losing visualization of a bunch of data. Here’s the problem:

• When a search result or link to my homepage is followed, MM doesn’t present it alongside my post pages (it’s buried on the most granular Link section interface)
• When a search result or link to a category/tag index page is followed, the same happens as above

Here’s a possible solution for keeping this simple and presenting the most data as possible:

1. Re-label the Posts section to become "Page Views" This basic nomenclature and data point isn’t represented anywhere in MM
2. On the first interface in the new Page View section, present the stats in one table and clearly mark each type of page view with a text or iconic descriptor. Then add a simple widget for choosing: All, Homepage, Categories or Posts.
3. Of course, make it smart to remember which view the user last used
4. Back on the homepage, bubble up the number of posts and categories viewed (out of how many exist) within the large icon, directly under the total number of page views.

Now, at a glance, I’d be able to see my total page views, while also being able to dig deeper and get a sense of which pages are being hit. Simple and powerful.

The United States Of America
I fully realize that the Country section is icing on the average analytic cake, but it is so much more than that within the context of a global perspective. So I’m thrilled to have a tool that visualizes for me where my visitors reside around the globe, while providing a fun geography refresher course. But now I want sweeter icing; I want to know where my visitors are coming from within America.

ESPN.com generates this very view when they present poll returns from around the nation. Yes, this is US-centric, and doesn’t provide a peek into granular levels elsewhere around the world, but if the data is available (which it is) expose it. The zoom feature practically begs for it to be implemented, as I’m dying to see if a rancher in Montana is connecting*the*dots.

I’m really looking forward to the beta release.

Back in 1999, I found myself living in the northwestern corner of Massachusetts, working in an area tagged as Silicon Village. Yeah, it was a little premature, just like it’s cousin Silicon Forest in Portland and it’s big brother, Silicon Alley in NYC, but the dot-com era was booming and the entrepreneurial spirit had caught both Williamstown and North Adams square in the heart.

(originally uploaded by Original Sin)

It was an exciting time.

Tripod (the company I joined) had just been purchased by Lycos (or as the long-timers liked to refer to them; the Death Star) for more than $50 million dollars. The young, personal web site building company and online community had made it to the big time; now one more trophy brand in Bob Davis‘ portal empire.

But Tripod didn’t start off as a personal publishing website; they flicked on the converted cable factory lights with the intent to provide advice for college students and post-graduates in print and on the internet, while the resume engine and online community all came later. DeWitt Clinton, a Williams student and Tripod programming intern in 1996 tells it like this:

“In the beginning — and this tension carried on for years — Tripod was a content company that just happened to use the Internet. (Recall that they also had a magazine and a book published.) Thanks to some clever people like Jeff Vander Clute and Nate Kurz, a few useful ‘Tools For Life’ such as the Resume Builder, were built. These applications were an interesting synthesis of ideas from the designer(s), editors and programmers.

I would definitely say that Bo was in a position of watching what everyone came up with, rather than intentionally leading them there, saying as much in his recent book. The homepage builder was just one of these organic and surprising inventions.”

So what happened? How did the tipping point occur within Tripod itself? When did management decide to move forward and focus on personal publishing and online communities? DeWitt adds more color from an outside, post-acquisition perspective:

“The traffic generated by the home pages earned them an acquisition, not the editorial content. See the Geocities acquisition just a few months later for evidence.”

Bo Peabody, Dick Sabot and Ethan Zuckerman hired some super smart developers to get their original concepts online. They built the first online resume engine and created a place for community to form the first iteration of Tripod.com. But a crazy thing was happening — people weren’t using their product the way they had envisioned. People were more intent on building their own web pages with the resume builder.

Damn these people!

Bo and company had a choice to make; either stick to their origin vision or evolve to support the needs and desires of their members, moving Tripod towards focusing on homepage building tools.

They made the only choice they could.

In 1997, before revenue models other than advertising came to fruition, stickiness determined the value of most companies. Bo and Dick saw the synthesis of member desires and a business opportunity, usefulness and viability.

It was a no brainer.

The Lycos Years

By the time I came on board, Bob Davis had just scooped up Tripod and Bo was serving his commitment to Lycos, wandering the halls at odd hours.

“Corporate refocus” was quickening its pace.

The first driver I encountered was the order to cut out community all together and focus solely on developing a suite of personal publishing tools. Actually, that became the name of our group within the Lycos domain: Personal Publishing.

The move ostracized many of the original Tripod folk who had joined the company because of the possibilities of online community, as well as a bunch of members that chose Tripod for similar reasons. But the numbers proved that people wanted to build their own web sites, so the machine spit out its orders and rolled on.

My first project was to visualize the current experience in a tangible format, so we could determine where we were going to snip and cut sections and features. After putting together a precise map of page sequencing and explicit sections, I walked into the office of my design director (former Tripod lead designer, Dave Reid) to get his opinion. The direction given to him was crystal clear, so he studied my map for a few seconds, found where the “build” and “community” sections bordered one another, and proceeded to literally rip the map in half on that line.

No questions asked; no questions necessary. That was how the Death Star operated.

It took me about a year into my stint at Tripod/Lycos to really start to question the direction of the group. I mean, the projects I was being assigned to were superfluous at best, such as creating Hello Kitty skins for the Angelfire publishing tool UI.

It felt like the powers-that-be had run out of useful ideas, so they just wanted their paid bodies in motion, any motion, as long as we were being productive.

That’s when I began wondering what would’ve happened at Tripod if they hadn’t been sold to Lycos; if the smart people were still “in charge,” still listening to their members, still innovating based on where they came from and an evolving vision of where to go.

Maybe Technorati would be serving the world of “Tripoders” today, rather than “Bloggers”…

As things would have it, Lycos closed up the Silicon Village web factory to prepare for the Terra merger. I wanted no part of working inside of the Death Star in Waltham, Mass, so I moved down to Brooklyn and picked up a dotcom consulting gig.

It wasn’t the best move, but it was better than Hello Kitty.

Jeff Veen’s post the other day regarding the genesis of flickr placed me in this Silicon Village time capsule. His description of their roots reminded me about choices and their consequences — good, bad or indifferent.

There’s no “right way” to create a viable, useful product; no methodology that is absolutely sound or fool proof. The best you can hope for — as Bo so eloquently points out in Lucky or Smart?– is that if you subjugate your ego often enough, and live your life accordingly, options will be presented to you in a manner that you can act upon with intelligence, vigor and respect.

That advise should be the first amendment for both creating useful products and collaborating with smart people, as in both cases, consistently relying on ones self-referential perspective is rarely ever a spot on decision.

Viva la flickr! Viva la Tripod! No game, no resume… no problem.

I’m now booked with reviewing hours and hours of user interviews, run by Ameritrade’s client experience consultants. I asked for it, and marketing gave it to me. The info is great — hearing the God’s honest truth about your site can be quite interesting — but the format of the interviews is pretty drawn out and somewhat unfocused. Unfortunately, we never get to see what screens they are reviewing at a given time because the brilliant camera operators don’t cut to a fire-wired view of the laptop screen. Details, details…

Technical problems aside, I’m moving forward with my design process as my initial three personae for the trading platform redesign project are being rounded out as the customers open up to the interviewer. Good stuff. Now I just have to roll up the specifics into a document that focuses on the goals of the individual personae and develop a few mental models for the beginning of the redesign of the site/applications.

It’s great to finally have the opportunity to do things the right way. Kinda.

We at Datek are only a few months away (hopefully) from signing on the dotted line and merging with Ameritrade. A major result of that milestone will include a re-architecture of the two discrete sites into one new user experience; a forward-thinking trading platform. In expectation of that date, our marketing group has been doing a great job in gathering user data (video tapes of user interaction with the two existing sites, focus group feedback, etc.) and we — the user experience design team — are about to embark on the first step of the the planning process; creating the first stab of an evolving bible of Ameritrade/Datek personas.

Before I started work on designing Command Center, personas were not part of the software design process at Datek. During the initial stage of the project, I contacted a friend of mine (CCO of Onclave, Dave Reid) to borrow a persona template. After mind-melding with Dave and then internally with our marketing department, I came up with two detailed personas two reflect our active trading customer base. Sure, they’re not based on personal interviews with clients, but as stakes in the ground, the tangible reference of human personas consistantly assisted me in the design of the product and proved to be worth the effort.

The shift of our software design process, from a business centered to user centered, has grabbed huge traction over the last few months, and I feeling that the change in approach will positively affect our customers interaction with the Winter ‘03 site launch.