Opening at Kellogg for a Web Design/Content Management/Content Editor

Our friends at Kellogg just passed this job opening on to us, check it out:

http://chicago.craigslist.org/chc/web/1539932196.html

Challenging job responsibilities, top-notch institution, and best yet, you’d be working with some really good folks!

Posted by Voltaire Santos Miran
Additional Posts (79)
Categories:

Discuss this article (0)

UB: mental models, unveiled!

Last week we unveiled our mental models to the University at Buffalo community—silent sigh of relief, big yay for all of us!

Four models in total:

1. Matchseekers: people evaluating whether an institution, job, or working relationship is a good fit for them.

2. IT Solution seekers: people looking for answers to IT-related questions or issues.

3. Health Solution seekers: people looking for answers to health-related questions or issues.

4. Active Supporters/Prideful Belongers/Pulsetakers: people who want to support the institution, as well as people who track specific issues of interest within the institution.

The work to-date represented over 1,000 collective hours of research and analysis. In our prep for presentations, Rebecca and I asked each other “So, what we learn, and was it worth it?”

First, the learning. My key takeaways:

1. We learned how to listen differently. The process of having these conversations talk us how to listen intently without leading or constraining. One person on Rebecca’s team told her that, as a result of this project, she’d never do interviews in the same way.

2. We confirmed some of what we thought we knew. Having done this for awhile, we thought he had a good sense of our audiences’ needs and expectations. In many cases, we affirmed that sense.

3. We filled in the blanks. We knew, for instance, that prospectives evaluated both the institution and the city in making their decisions. Now we know that they evaluate the city and institution by different criteria, and that criteria changes from mental space to mental space.

So was it worth it? Yes, indeed. Particularly for developing content—the mental models give us a good deal of information about framing and delivering information to better meet the needs and expectations of our target audiences. The models also help in developing detailed architecture—knowing more about the what, when, wheres, and whys helps us to create link sets and information clusters more effectively.

And the models scale. We’re applying the matchseeker model to sites for the medical school and to university communications as a start. That same model can be used by the rest of the schools at UB as the Web Content Initiative rolls out through the institution over time. And that model can be expanded, with additional research being combed into the model to continue to enhance our understanding. 

My favorite moment came when a faculty member from the medical school commented to me that the information he saw in the models confirmed his experience—not only as a doctor and researcher, but also as an individual. That affirmation, priceless.

The next few weeks, another bit of a race: baseline testing, information architecture development, usability, and wireframes…

Posted by Voltaire Santos Miran
Additional Posts (79)
Categories:

Discuss this article (0)

UB: The Race to Place a Mental Space

Don’t get me wrong, I love what I do. But I’ve quietly opined to close friends from time to irregular time how I miss my first career as an underwear model. Thanks to this UB project, I’m a model once more. A Mental Supermodel, that is. You’ll forgive the lack of an accompanying graphic (my salad days are over) as I explain….

Grouping began roughly four weeks ago. Roughly 50 interviews, each with roughly 60 nuggets each (3,000-ish nuggets total), to be sorted into towers and mental spaces that make up the mental models for our audiences.

Indi’s gentle guidance kept us on track through this part of the process. Notes from our training sessions with her:

• Avoid grouping by the nouns. The temptation to cluster tasks around topic is great: resist! You may have five nuggets that contain Buffalo, the city—they don’t necessarily belong together.
• Each voice should appear once and only once per task. This rule forces you to do two things: either to combine two or more nuggets that really represent one idea, or to tease out the subtleties that make each nugget unique.
• Suspect your combs. If you’re just starting out with the mental model framework, you’ll probably need to rewrite (or as Rebecca kindly puts it, “mature”) more than half of your nuggets. That’s OK.
• Realize that the beginning of this is the hardest part. The first few rounds of grouping can be really painfully long and disheartening (where’s the pattern, where’s the pattern?). Grouping goes faster later in the process.
• Expect the process to be fluid. Deconstructing and reconstructing are part of mental model development. You may touch or relocate each task up to 10 times or more as the mental model builds. That’s better than OK, that’s essential.

• Resist the need to find a place for every nugget you’re combing in. Some nuggets will become tasks or towers eventually—particularly if they represent a philosophy, motivation, feeling, or action that you think would be expressed or experienced by another individual in your audience segment. Sometimes, one can stand alone.
• Pay attention to when, as well as the what and why. In the matchseeker mental model, for instance, we have towers related to evaluating the city early in the process, during deliberations, and after decisions are made—matchseekers think and do different things related to researching a city as they move through the process. If you don’t pay attention to the timing, you can lose valuable information.
• Give yourself time to think and breathe. It’s like doing a crossword puzzle—walking away for a little bit allows your mind to process in the background … to recognize patterns and reach better conclusions. Indi told us to budget three hours on average for each conversation you group into the master—it took us a wee bit more, but we had fun doing it together.

We started grouping together as one large group, working virtually in half-day sessions to get a rhythm and develop some shared sensibilities. Then, to move things along, we divided up into two groups, the Honeycombers and the Mental Supermodels (guess which team I was on!). And for the last stretch, we broke into three teams to finish things up.

And at phase end, we’ve got four mental models: a matchseeker/pathfinder model, a pulsetaker/prideful belonger/active supporter model, a health solution seeker model, and an IT solution seeker model. And I must admit, I let out a sigh of quiet relief at the conclusion of building each diagram out (a process made infinitely easier by the drag-and-drop diagram capabilities of OmniOutliner and OmniGraffle). It’s sort of like being allowed to exhale while the photographer reloads.

Next week, previews of the new models to the information architecture team. Welcome (back) to the catwalk! 

Posted by Voltaire Santos Miran
Additional Posts (79)
Categories:

Discuss this article (0)

Oh, My Kingdom for a Finely Tuned Comb

Indi Young, Mental Models, page 164: “The pain you feel is weakness leaving your body.” Amen, sister!

You have to imagine the scene to fully appreciate it. Picture Jeremiah, Laurel, and me—each of us tethered to our laptops by our noise-canceling headphones. Listening, listening, listening. Pause. Transcribe. Listen. Back up the recording. Play. Didn’t back up enough—more, more! Play. Transcribe. And, nugget. Then imagine our UB teammate, Eileen, doing much the same in her semi-private office space (or, for short, SPO) in Buffalo. 

The process is called combing—reviewing the audio recordings each of our 50-some conversation and extracting root tasks (or what we call nuggets) that will be later grouped together—from atomic tasks into tasks, tasks eventually clustered into towers, and towers eventually clustered into mental spaces. Each hour-long conversation requires roughly three hours of analysis and yields somewhere between 50-75 nuggets.

It was, honestly, the most painful part of the process to-date. Now that we’re in grouping mode, I’m really gratified to see how the hard work we did together is paying off. But before I talk more about what we’re finding in our grouping (next blog post, I’m such a tease), some notes from this part of the journey that I compiled in talking with the team:

1.  Ideally, you should have both audio and electronic transcripts. Time and cost considerations moved us to an audio-only combing process. We did a few proof-of-concept combs this way, thought it worked out fine. But what we didn’t take into account is how difficult it becomes to comb this way when you’re combing three interviews (or, nine hours’ worth) in a day. In one instance, I had both the original audio file and the electronic transcript for an interview—I was able to finish that comb in roughly two hours (a 33% time savings) and ended up with more nuggets than I had with audio-only combs. I think we ended up with quality material overall—but combers will have a much easier time if they’ve got both types of files at the ready. 

2. It’s also easier, when the comber is the same one who did the interview. We’re talking ideal, here—the person with the skills to lead the interview isn’t always the best person to comb out the details, but I found that it helped me a great deal when I combed my own interview.

3. Barring that, it’s best to read through or listen to an interview in its entirety before you start to comb. This will give you a better sense of the conversation as a whole, and will also help you to find the right nuggets and connect the dots for nuggets that must be pieced together from different parts of a conversation.

4. When in doubt, capture the quote, then come back to it later. Sometimes, you know that the quote was significant, but the right label for that nugget will elude you. Fear not—note the quote and write the label later. It’s like doing the Times crossword—put it down for a bit, and suddenly, the answer comes. 

5. Use their words. In writing labels, you should use the words that they used. The temptation is at this stage is to truncate a phrase or use a word that’s more succinct in your mind. Resist. Using their words will help you to remember the conversation and context, which is essential when you’re  trying to spot similar nuggets in a sea of 1,10o of them. 

6. Except when you shouldn’t. When shouldn’t you? When the term only make sense to you as the person who led or combed the conversation. One person I interviewed talked about getting out of the town’s “bubble.” In my mind, the entire riff on Stepford Wives that accompanied that comment was fresh and hysterical, but none of the other combers was in on the joke. I changed the label to explain the “bubble.”

7. Know the difference between feeling and believing. As Indi explained to me, a belief is a core value upon which individuals make their choices. A lot of times what sounds like a belief is really a feeling or informed opinion and should be labeled as such. In explaining this to Laurel, I used the following example: someone very close to me doesn’t believe in life after death. He believes that his time on this earth is what he’s got, and he’s intent on living the best and richest life he can because there’s no “to be continued.” He also likes wearing golf shirts because he believes they’re comfortable and slimming. Only the first of the nuggets qualifies as a real belief.  

8. Have the right tools for the work. If you’ve been looking for an argument past OSHA for a chair with back support, a large monitor or laptop stand, the new SONY digital noise-canceling headphones, or an ergonomically positioned keyboard tray, you’ve found it. 

And, finally, a quote from Miah that I believe represents us all:

“I wish I had seen the need to slow myself down and just immerse myself in this earlier.  I’m so accustomed to having email, IMs, web browsers, office phones and cell phones all occupy a portion of my attention all of the time, that it was really strange to be working on something that could not allow for any outside distractions … after getting over the initial anxiety of being removed from my “lifelines,” I was able to relax and understand the time that the combs were going to take, and to not get anxious that I was missing something.  The world went on without me, go figure.” 

I count six nuggets, anyone else?

Posted by Voltaire Santos Miran
Additional Posts (79)
Categories:

Discuss this article (1)

UB: Week 12, Unlearning My Cocktail Conversation

“Thank you. You helped to realize something about myself that I didn’t know until now.”
“No, no, thank you!”

One of the best conversations I had during the interview stage of this project was with someone who was talking about the people who made an impact on that person’s choice of career. It was fantastic, and according to Indi, exactly the type of conversation we hope to have with people.

The last few weeks, a big push to complete our calls—over 47 hour(ish)-long conversations, in total. The majority of them were fantastic, and while we’ve got a few left to go, we’ve started the combing process. What’s combing? It’s basically sifting through each of the conversations. We’re searching for nuggets—compact phrases that use people’s specific words—that identify behaviors, feelings, philosophies, intents, and all of the motivations that drive them. Check out Indi’s blog about this—and just to be clear, Rebecca anchored to the panning-for-gold analogy, while I went for the more pedestrian Mickey D’s. reference.

But combing is a post for another time. Today, I wanted simply to share my personal lessons learned about interviewing for mental models.

This process is seriously different than talking with stakeholders. In stakeholder meetings, you’re doing triple duty. Surely, a great deal of those meetings is devoted to finding out what stakeholders know about the audiences that they’re trying to serve. But part of those meetings is also spent building the client relationships—client and consultant need to get to know each other and to trust in each other’s expertise and experience. And part of those meetings is devoted to planning—talking about the entire process, next steps, etc. Focus groups with target audiences get us closer to source information.

These conversations, closer still. We’ve been talking directly with individuals who represent the audience segments that we’ve identified in our mental model (and as quick reminder, we’ve got six: matchseekers, pathfinders, solution seekers, pulsetakers, prideful belongers, and active supporters). Indi likens these conversations to ones that you might have over cocktails or dinner, and she offers the following rules (which you can read more about in Mental Models):

1. Behaviors and philosophies, not product preferences
2. Open questions only
3. No words of your own
4. Follow the conversation
5. Not about tools
6. Immediate experience

Only, I realized that I needed to unlearn my cocktail conversation. So let me offer my own few add-ons:

1. Don’t worry about finding common ground. You know what that’s like. You meet a new person at party and you begin to talk and the first thing you want to do is find something that you have in common—a topic you can banter about and discuss. Something that interests you, and on which you have opinions and thoughts that you’d like to share. Except that this conversation is solely about the person you’re talking to. Which to an introvert like me is a bit of a blessed relief.

2. Don’t worry about being interesting. Or sounding smart. You know what that’s like as well—oh, the pressure to be witty and engaging and loved by all! Except that, again, it’s about the person you’re talking to. Their impressions of you only matter inasmuch as they can sense that you’re truly listening, paying attention, and grateful for the time and disclosure that they’re willing to give you.

3. Don’t avoid the childish question. “Why?” “Why?” “Tell me a story about …” As long as the person you’re talking with is forthcoming, the simplest questions get the most interesting and fertile answers.

We’ve got 47-some conversations to parse—an exercise that will probably yield almost 3,000 nuggets, all in our target audience members’ own words. Then, grouping. User-centered development, anyone?

Posted by Voltaire Santos Miran
Additional Posts (79)
Categories:

Discuss this article (1)

UB: report from the front, week nine

I believe that Amy Grant’s music, much like rose wine, is much maligned. There’s a line in one of her songs, for instance “… how do you argue with a feeling in your bones about what is and what isn’t meant to be.” Makes me think of the UB project.

Let me explain, and sorry—by the way—for not keeping up with my posts on this initiative, it’s been a bit of a whirlwind set of weeks.

Last I reported, we were beginning interviews for our mental models. Five different audience segments, 70 individual conversations. Three interviews in, both Rebecca and I sensed that something was amiss. We couldn’t quite articulate what wasn’t working, but we knew that we weren’t getting the information we needed. Part of the issue, we thought, was not having the right people to interview. But there was more to the issue than that, we knew. She and I talked, decided to sleep on it; when we met the next day, we found that we’d arrived at the same solution by different routes. The answer: re-engage mental models mastermind Indi Young to help us revise our scope statements, re-write our prompts, and lead one model interview for each of the audience segments we were studying. And that’s exactly what she did, and it’s exactly what we needed. (And Indi, if you’re reading this, you’re my hero!)

Toggling like this cost us some effort—it meant rescreening and rescheduling a number of people. It cost us money. It cost us time. And it also cost us some comfort—I had to learn a different and new way of interviewing. But it was exactly the right thing to do for the project, and it also led to some really terrific outcomes:

1. We learned that we didn’t need to interview as many people as we originally intended. In fact, we were able to cut the list by over 30%.
2. We found a new audience segment—people we call pathfinders, that fall between matchseekers and solution seekers.
3. We found that our six audience segments divide into two clusters, or clouds, or continuums, if you will—each cloud having a number of shared characteristics.

The interviews are rocking and rolling now (go, UB scheduling team!); we’re slated to finish them up in the next week or so. And in the meantime, we’re going to start combing through each of the completed interviews for behaviors, feelings, philosophies, intents, and motivations. How fun is that!

Posted by Voltaire Santos Miran
Additional Posts (79)
Categories:

Discuss this article (2)

The UB Project Report: Week Five

End of week five, Buffalo. Last week, we completed a fairly intense daylong workshop with Indi Young and Eric Fain, devoted mostly to the principles of non-leading interviews and the process of combing through transcripts for the tasks that will eventually be the building blocks for our mental models.

I’ve come to two conclusions:

First, the best interviewers for this process are the ones who have:

a) gone through several years of therapy and/or have a parent in the counseling profession
b) done a good deal of teaching in some sort of setting
c) led an awful lot of intake meetings over time

As Jeremiah put it, it’s a hard thing to pull back from “tell me what you want and need, oh my wonderful client” to allowing the conversation to go where it needs to go. Indi likens it in her book to the kind of conversation you’d likely have with someone at a cocktail party. (Funny, but those are the hardest to remember through the martini haze.)

The second conclusion is that parsing human behavior is a hard, hard, hard thing to do. One of the things that I really love about Indi’s mental models framework is that it provides a fairly granular and scientific methodology for understanding your users’ expectations and needs. It’s time- and labor-intensive (and it also requires a great deal of discipline in consistently dissecting from interview to interview), but it’s thorough and the resulting IA is imminently defensible.

Our project team spent most of this week laying the technical groundwork for scheduling, recording, and processing phone interviews—Google Apps, Skype and plugins, Garageband, and Quicktime, oh my. Next up, 66 one-hour calls over three weeks. We’ve got four interviewers and four audio-transcript combers prepping for the breakneck effort. Let the wild rumpus start!

Posted by Voltaire Santos Miran
Additional Posts (79)
Categories:

Discuss this article (0)

All the Pretty Little Rows and Columns

I will resist doing it my way, I will resist doing it my way, I will resist doing it my way.

Week three of the Buffalo project (see my intro post), and things have gotten really interesting. I was telling Rebecca yesterday that I had originally wanted to develop the information architecture offline using mStoner’s standard process, just to see how similar it would be to the structures we’re developing using Indi Young’s mental models process. But I decided that the temptation (conscious or subconscious) would be too great to match the results of our work together to my super-secret preliminary efforts.

So we list and group and nest and cluster and play Tetris, just a little bit.

Let me explain.

Normally when I create information architecture, I start with my clients’ primary target audiences: usually, prospective students and their influencers, and prospective donors. For each of those audiences, I list key tasks and critical paths to accomplishing those tasks, and I flex the site structure to make finding that information and completing those tasks as quick and intuitive as possible. Harder than it sounds, but that’s a post for a different occasion.

With this process, we dial back even further, and we use a spreadsheet (or in my case, an Omnioutliner file) to:

1. Create a laundry list, row by row, of all the tasks (which include general information and services) that people would want to accomplish on the UB site. We then sort and group those tasks to eliminate redundancies.
2. List the different people who might perform these tasks in different columns.
3. Mark x’s in the cells for performers who might complete those tasks.
4. Visually move the columns and rows around until we start to see visual patterns and clusters (see, just like Tetris).
5. Name those clusters with descriptive labels.

And at the end of the exercise, we don’t have prospective students, current faculty, alumni, per se. Instead, we’ve found match seekers, solution seekers, prideful belongers, pulsetakers, and active supporters. Rebecca shopped these groupings around to some internal stakeholders, and eureka!, they make sense.

Next up, training next week with Indi and her colleague Eric, and then we begin a slew of one-on-one interviews to validate out hypotheses about their needs and expectations and to flesh out the task list. Stay tuned, same bat time, same bat channel. 

Posted by Voltaire Santos Miran
Additional Posts (79)
Categories:

Discuss this article (0)

I Get to Be Batman

I love all of the work we do, but I’m especially stoked about our newest new business. Earlier this week we kicked off a project with Rebecca Bernstein and her team at University at Buffalo. An aggressive, five-month timeline of what I call extreme IA, as we develop, test, and refine structures for eight websites. It’s a special treat for me in three ways:

First, the project gives us the opportunity to really explore how the needs, expectations, and behaviors of site visitors have changed over the years. I’ve been producing university sites way before mStoner was founded in 2001 (Mosaic and NetObjects Fusion, anyone?), and I’ve learned a lot. This project gives me the opportunity to validate some of what I’ve come to expect, but also hopefully to flex and adapt my understanding to new trends and possibilities.

Secondly, this work allows me to use a new way to develop information architecture. We’ll be using Indi Young’s Mental Models as our framework—I’ve been a fan of her work for years, and this project provides time and budget to do justice to Indi’s process.

Finally, the expert-on-expert factor. I don’t know anyone who’s won more awards for eWork in higher education than Rebecca, and she’s got an incredibly bright, talented team by her side at UB. The chance to come to the table with them and share ideas, debate, discuss, push, pull and hopefully arrive at moments of collective brilliance … very cool. Rebecca and I were talked yesterday, and she commented that it didn’t feel like she’d hired a vendor—it was more like she expanded her team. Or, to riff on our lunchtime discussion during immersion earlier this week, it’s like the Justice League of America. Each of us a superhero in our own right (or own mind), bringing a special skill, talent, and approach to the task-at-hand.

I get to be Batman. Why? Consider the character: incredibly wealthy and handsome man, strength and brawn and fighting skills beyond belief and bolstered by cool gadgets and immense technological resources, haunted by a tragic past, only slightly to the right of the villains he pursues, and looks great in form-fitting latex and a cape. The parallels between us are staggering.

We’ll be sharing what we learn by blog, so stay tuned for some exciting stuff. Same bat time, same bat channel …

Posted by Voltaire Santos Miran
Additional Posts (79)
Categories: Strategy

Discuss this article (0)

Respecting the Site Visitor and the Medium

This week, workshop madness! I just completed a series of four of them for a local client—they had selected four pilot units within the college to go live with new sites for their specific areas at the same time that the main, public site launches later this year.  In a three-hour format, we reviewed the principles of visitor-centered design, information architecture, writing for the web, and navigation development and then wrapped up with a chalkboarding exercise in which we started to wireframe the new homepage for that unit, based on principles that we’ve discussed earlier in the day.

The goal: to help pilot units understand how and why we’ve created the information architecture and navigation for the main college site and to help them flex that model to their specific needs.  In one workshop, someone asked me if the goal was to design to the lowest common denominator. In another, one participant asked if the goal was to design for a 20-year-old (the reason he asked was that he, as a seasoned academic, had come to expect text-dense, very long, and formally worded prose). Both questions really caught my attention and made me think. And in both cases, the answer was "no, not really." I personally think that the goal is to serve your site visitors well by respecting both them and the medium.

Steve Krug’s book "Don’t Make Me Think" was published in 2005 (eons ago, in the web world), but it remains a touchstone for me in all of the planning and training that we do. Its main premise: that websites should be so intuitive that people don’t need to question "where will I find this" or "how do I accomplish that?"

Business school taught me to think in threes—the top three principles I cover in our workshops:  1. Sites should be designed with your site visitors in mind. And those site visitors don’t think in terms of organizational charts or industry jargon (in one workshop, someone explained to me that course articulation translated into "will I get credit for this course;" who knew?). 2. Some site visitors self-identify. Some wayfind from topic to topic and link to link. Some search. Some think in terms of tasks or "I want tos…" Most will do all four, depending on the information they’re looking for.  3. People skim pages more often than they read.  When you based site design on these principles you inevitably gravitate toward labels that are simple and straightforward and clear. You also offer multiple entry points to accommodate the different mental models that people use for parsing information. And you write with ruthless journalistic discipline—being as compelling and concise as possible.

I’d argue that sites designed in this way actually broaden your reach and appeal, and that’s a good thing.

Posted by Voltaire Santos Miran
Additional Posts (79)
Categories: Content and writing

Discuss this article (1)