Dan Michaelson talks with John Caserta
Dan Michaelson is a co-partner in Linked by Air. John was joined by Catherine Schmidt, Evan Brooks and Michael McDermott for the interview on February 22nd, 2014.

In this class, we are dealing with certain areas of the browser as a tool, particularly as a production tool, severing off the consumption or interactive side. What is the browser capable of? You can make things in it. You can also receive things in it. What is the browser? You work so much in it and your work seems mostly browser-based in some capacity, whether it’s serving someone a browser deliverable or not. How do you see it or use it in this day and age?

In our studio, and definitely in my class which is called Networks and Transactions I’m almost focused on the opposite actually, on the idea of the browser as the end of a network that’s reaching out, and on HTML as not only a formatting language but also a HTTPS protocol that really is all about distribution. But I think your class is great and I think that that very distinction suggests the need for filling out the other side of that equation as well. I think browsers are rendering engines and they’re increasingly powerful ones for sure. More and more you can make anything in HTML and CSS, especially if you’re not worried about legacy browsers and stuff like that. I also think that the inspector is an incredible tool for manipulating a design language and an interactive language. Are we talking about the browser just as a rendering engine, or are we also talking about the browser as an interactive platform, even if we’re still ignoring the network? You can tinker with the code behind the rendering or with the code behind the interaction, even while the interaction is actually going on. That itself is a paradigm that doesn’t hardly cease to exist anywhere else. You can’t do that with a car. You can’t open up the engine and put the fan belt on a different flywheel while it’s still moving, but with a browser inspector you can. That’s an amazing way to think about design: as a machine, but one that you can manipulate in real time.

That taps into another question that Professor Lucy Hitchcock said after your lecture, “there’s really something happening in our field where there are the formmakers, whether its graphic form or what have you, that illustrate concepts and ideas, and there’s the systems makers.” That split struck me because, on one level maybe that’s true, maybe that’s obvious. Could you see our field as made up by those camps, and if so when did that begin? I think when you were a grad student around 2000, you probably have seen much of that split occur if that is indeed the case.

I think that’s a useful distinction. I try to find a way to talk about systems as form, so if we’re trying to predict the way that people move through a space that movement in us is also a form. I think graphic designers have always been comfortable talking about things like rhythm and color. It’s not a problem to think about movement as a kind of form in, for example, a color theory class, or when discussing Saul Bass film titles or something. If you are talking about the speed or the qualities of how people move through a building that you’ve done the wayfinding program for, it seems like that’s an aspect of the form of that wayfinding system. Similarly, if we think about the rhythms or qualities of how a person navigates their way from section to section of a website, that’s also a kind of form, just the same way you might think about the rhythm of the hierarchy of a complicated instruction manual as a kind of form.

And it requires the user to have that play out you think? Or is it actually inherently there.

Yeah, I think that’s a good point. I don’t think it’s new honestly. It requires the user to play out, but isn’t that true if you think about a reader as well, and then ultimately when you’re designing typographic hierarchy? I think we got used to thinking about those things as being things in themselves, and yes I can see the typographic hierarchy as an image on my retina, but actually it’s always been about the reader. I think there were some really great systems thinkers in the 50s, too, and I wonder whether for awhile we forgot about that and started to think of these beautiful finished compositions. The signifiers of “good design” took on a life of their own. Form is always something that is serving a function, and function is something that is completed by a reader or consumer or user, and we think to the experience of those things always as a part of the design. It’s not to contradict Lucy’s point at all, but just to say I don’t think it’s entirely new, but it’s an important aspect about how to think about Graphic Design. Also, I would try not to describe it as a formless idea (some people maybe would, and it may be useful to talk about drawing a distinction between content and form and experience and form) but I actually think it’s nice to think about what are the formal qualities of an experience.

I’ve been looking at the Graphic Design Now in Production catalogue and Andrew and Ellen and Jessica Helfand are coming to do a panel in a month or two. The show’s coming here, and I think its been 3 years since the show went up, and it was probably curated 4 or 5 years ago. Most of it is graphic form/ graphic art. There is something to chew on. The Lust Poster Wall is there because it produced a wall full of digital posters and so I wonder, were you comfortable with the way that show was curated or worded, even the sections like “information design,” and “posters.” If so then what are your thoughts on it now as far as how it describes the profession, because Yale still describes itself as there is Graphic Design within the School of Art, so you share that term with the Walker, and the Cooper-Hewitt in a sense.

I don’t mind the term Graphic Design at all. It’s such a nebulous term. Like I said, I think it’s a pretty open-ended term to describe ourselves that way. I think that visual form is really important too, and people might disagree about whether our work is beautiful or not, but it ought to be. We’re trying to make beautiful work, and we’re trying to make beautiful outcomes, even if that work is sometimes completed by other constituents. I don’t necessarily think that the focus on two dimensional form is misplaced. You can imagine in a more progressive way to categorize work that isn’t so much about genres of finished work, but is about genres of process, or genres of functionalities, or different kinds of systems. Is it content management systems and real-time applications? That’s a funny distinction between functionality and visual design that we get caught up in sometimes. It is hard for us to capture the beauty of our work in a single screen grab, and that’s something that we struggle with in a practical way just working on a website. Right now we decided we should, on the homepage of our website, be a little bit more marketing focused and feature 4 case studies. We just made it quickly, but right now these 4 studies lead with just a screen grab of the homepage, which is not particularly compelling. I think those websites are very compelling websites, but it becomes more compelling as the series of screens accumulates through your experience moving through the site, even just it’s more compelling to scroll down than it is to just see what’s above the fold for example. So to address all those questions as to how do you distill an experience into an image, I don’t think its a question of whether that experience is formally beautiful or not. I would argue that it is. As visual design, how you take a crop of it, either through time or through space, is tough. To your point, it’s just a fact of communication and our salesmanship that you do need to find ways to summarize, give an elevator pitch, [etc.] There should be a compelling screen grab that can tell the story of whats good about a website, and maybe that’s also part of being a good designer is that it has that level as well.

How does Zaron play into this, the all-seeing eye?

Its a reaction to that, in that if we could just capture every single state of every website that we ever made, then that would be a way to more accurately represent what we’ve done as it has unfolded over years. Another cool feature of Zaron is that it can capture what I call “comic strips” of an end-user’s use of a site, so if you are a web designer, or if you’re organizing a urban game like Midnight Madness, the tragedy is that you never know what people are actually doing, what the users or players are actually doing with your thing, and that’s what its all about. If you make a poster you pretty much know or you don’t care looking at it what their experience is, but it’s still fun to go out to the bus shelter and see people looking at your poster, but theres nothing surprising. It’s just gratifying. With a website you want to know the paths people are taking, or with an urban game you want to know how people are stumped, how they are having fun, which way are they going and how are they using it, but you can’t be there. In later iterations of Midnight Madness, the urban game, technology has advanced enough that we actually could be outside organizing the game in real time, so that did allow us to get some cross-sections or glimpses through the progression of the game. But in earlier years, we would all take turns. We would be holed up in our game headquarters, and if there was a part of the game that broke everyone would be dying to be the one that would go out and fix it because it would give you the chance to cruse by and see players doing their thing. Once we had this mobile technology and we could have our game headquarters be in a park, the first thing that happened was an enterprising team stole our packet of clues. They said once we were actually among the users that, of course, the users can mess you up. We became more rhizomatic and flatter, but that’s also with the corresponding loss of security. It became a pure network with all the plusses and minuses of a system like that. So getting back to Zaron one thing we do know the sequence. We can’t really spy on people’s screens, and it’s not really possible in real time to know what people see on their screens, but we can spy on the logs of our server, which is a centralized thing but we do. You can track a user session and say the user number, which can be an IP address, but usually it’s just a long hash of their session ID which is just an anonymous random number. We know that they viewed this URL followed by this URL followed by this URL followed by this URL, and then what Zaron can do is render a screen capture, not of their screen which is not really possible, but of Zaron own screen but of that URL. What that produces is a comic strip of thumbnails of all the pages that one user viewed in sequence. Then you can see a series of those comic strips of many users, the sequences of pages that they saw. That becomes a pretty neat way of visualizing how a lot of people use the site more or less at the same time. Even moreso than looking at analytics or something like that which shows that as a chart. Being able to see that (even though its a simulation it’s something close to what they saw) is a pretty neat situational awareness.

So you’ll see the exact image that was there?

It can differ a little bit because we’re not using their browser and for various other reasons.

There certainly is a trend on using user data, probably in a different way then what you’re talking about, analytics or what have you, to make design decisions. Obviously when you take that to the extreme it seems nuts, but you’re doing that in a more thoughtful way. How does that influence the iterations that you make and how it plays into things like risk taking, client expectations and then advancements in the browser and technology, where maybe you can do something else now so all that knowledge doesn’t apply?

Our usual approach is to make a guess about how users would behave. Also, I think it’s important to make a stipulation as to how you want users to behave, and not to just be driven entirely by what users think they want, or what they already know that they want based on their experience to date. I think it is really crucial that design be progressive. You are making the feature so you design for that, but then you look and see if it worked. If the uptake of your new feature isn’t what you wanted relevant to the importance of that feature then you change it.

Does it actually match the hierarchy? So do you want 20% of viewers to actually be playing with this thing? Do you have an image in your mind? If you take the Whitney for instance, everyone should know the Biennial and should hopefully engage with that content in some way, but it’s likely that 30-50% would want to know something like visiting information. How do you know if the predictions were right? To what extent is it that the design’s fault or success versus what was going to happen anyways?

If you are talking about, lets say the Biennial compared to the opening hours, or even something like the Whitney’s permanent collection online compared to the opening hours, I think there are different kinds of information that need to be conveyed. We would assume that a lot of people wanted the opening hours, so we would go into it with that baseline assumption, even thought that’s not the most exciting or interesting thing about the museum. We make assumptions about what the website is for, and one of the things we know that its for is for people to find out the address of the museum, so that goes right up at the top. Maybe its not big, but it’s at the top. There are different ways of making hierarchy, and with something like the collection or the Biennial, you look at what percentage of viewers are clicking or landing in that area. If it’s low and you know that area of the website is a) important to the institution and b) actually well developed on the site, which is another factor. There’s a lot of museums where the collection is really important but is kinda crap on the website, and you shouldn’t expect a lot of people to be there. Ideally, in that case that fact that they’re not clicking through ought to drive development of the content. In that case it shouldn’t drive you to keep making the link bigger and bigger. But, if the content is great already and people aren’t getting there, then it does drive you to make the link bigger and bigger in some sense.

I had a question about the word “browser”: You talk about the word browser as designing with the browser or designing for the browser? Are we talking about browsers, or are we talking about HTML as a design language?

We are talking about the browser itself as a rendering engine… Its ability to accept server data, its ability to compute on the fly, your ability as the maker/ designer. You’re interrogating it as a maker and not an end user. Cutting out the end user, maybe that’s where the line is. What can we do with this as a replacement? So HTML and CSS become defacto base to what the browser wants.

You’re not literally designing in the inspector (that actually kind of a neat idea), you have a text editor and you’ve got this other tool as well that then is rendered by the browser …

More Conversations

Peter Bil'ak talks with Catherine Schmidt

Andrew LeClair talks with Christina Rees

Jason Pamental talks with Ojus Doshi

Jürg Lehni talks with Philippe Cao & Daniel Giuditta

Ben Pieratt talks with Lukas WinklerPrins

Dan Michaelson talks with John Caserta

Paul Soulellis talks with Christina Webb

Augustina Bello-Decurnex talks with Kelly Walters

Armand Mevis talks with John Caserta & Dylan Fracareta