In Part 1 and Part 2 of this series, I identified the elements that comprise an interaction model: UX patterns, feature clusters, system behavior over time, and UX principles. What sews those elements together is what I had been calling an "interaction metaphor," or the metaphor of the system.
So, how is an interaction metaphor different from the parti, as described in the previous post—the central idea or concept of a system? I'll venture a guess, though this is still rough thinking and confined just to interaction design.
The parti is an idea or concept describing how an interactive system should be constructed, from the top down. It aids designers in organizing and shaping the design around an optimal flow for users. The parti from Luke Wroblewski's Yahoo home page redesign: "The dashboard for what you love on the Web."
The interaction metaphor describes the spirit of how a system should behave, from the bottom-up. It identifies the core meaning a system provides emotionally, through its actions in concert with the user. In the case of the Yahoo redesign Luke showed us at Interaction 09, an interaction metaphor could be "Dashboard." However, a stronger metaphor could describe how the dashboard behaves more explicitly in use. In playing with Yahoo.com's current home page design, the metaphor that initially comes to mind is "Windowpanes."
If you've created a really strong parti, it probably has an interaction metaphor within it—but it isn't required for the parti to help you in organizing a system. As an example: In 2009, I was working with a team to redesign a website for a popular yearly event. We determined through the discovery process that site visitors were yearning for easier access to content throughout the year, not only during and immediately after the event. (The assertion was rooted in site metrics, event survey results, and other modes of direct feedback from a passionate user base.) We crafted a parti as part of the experience brief and used that organizing principle as a method of streamlining access to that content. The parti supported some major architectural decisions, such as stable year-to-year accounts (for regular access), which was then passed along to registration partners. There was no deep metaphor embedded in the parti—it boiled down to the idea of "drinking from the fire hose." The actual interaction design was geared around removing any mechanisms that stood in the way of that primary user goal.
In the work that I've been doing over the past year, however, setting a provisional parti into place and watching it evolve has caused unique challenges. It's been my observation that the greater the level of complexity of a system, and the less definition there is about what requirements exist for what the system should provide, a parti isn't terribly useful until you are very deep into the design work. When brainstorming for interaction, it can be almost impossible to craft a parti or an ideation metaphor from the top down, rather than working from the bottom up via features, patterns, and discrete flows. Defining the overall metaphor of the system happens once a working model for the entire system is in place, but has yet to be proven out through full implementation.
Crafting that metaphor or series of metaphors, and really understanding its impact on the patterns and principles that govern a system's construction, is almost the reverse of "big idea" brainstorming and seeking ideas that have "legs." By this, I mean that you can spend a great deal of time imagining up all sorts of grandiose metaphors, but none of them will work until you have the necessary material to prove it out. You end up discarding hypothesis after hypothesis, directly focusing on what's up on the wall, in the sketchbook, prototyped in code, and so forth.
In graphic design, the "big idea" is the grail of any brainstorming effort. This big idea has "legs" that can be infused via aesthetic character and patterns of expression across a range of deliverables. Well-executed visual design systems weave that metaphor clearly and consistently through all designed communications vehicles, usually driven by some select set of drivers or an insight statement.
A "big idea" can resonate with your gut, and move you. Usually, these ideas are supported by brand or messaging pillars, which are rational and emotional support statements meant to resonate with or influence the thinking of anyone who comes into contact with those designed elements.
In contrast, an interaction metaphor can start as a tangible image, feeling, or even a mythical concept. However, as the interaction metaphor plays out in the formal design—or, more likely, is back-rationalized and infused into a system beginning to cohere—it begins to operate on multiple levels. It infuses the patterns, principles, system behavior, and feature prioritization.
The holy grail for any product or service designer is to be able to elegantly describe the metaphor of the system, then prove it out through usage scenarios. In some design presentations, it can take a single word or phrase to eloquently connect key moments in a system's information architecture, interaction design, and visual design. When you encounter a system design with a compelling, unifying interaction metaphor, you can often detect that metaphor through the "iceberg tips" of wireframes, prototypes, or animatics. And when users interact with the system in testing, they often react to how those details weave together into a (hopefully) cohesive whole.
Some Approaches to Identifying an Interaction Metaphor
All right: I want to create an interaction model, and by extension, nail the metaphor that holds the system together for users. How does this happen as part of our working process as designers?
While we may not be able to explicitly come up with a discrete idea that immediately organizes a group of other ideas, we often discern from the design material the shape of an interaction metaphor. It emerges from structuring and synthesizing ideas post-brainstorm. Write words and phrases around clumps of design ideas that describe how they may behave in implementation. In previous projects, we've taped up on the wall or pinned to black boards zones for sketches. We created clusters for specific features, then began to abstract from those features repetitive patterns. We also outlined on sheets of paper principles as to how elements might be structured, or rules as to how many actions are required to reach content. We also created ideas that outline things the system might say, the tone of voice that it uses, ideas for natural movements and gestures, and so forth. (If we're feeling spry, we'll spend a few minutes considering what kinds of metaphors in sum may govern the overall system. This could be as simple as bandying about concepts such as Nautilus, Campfire, or Unicorn—though Unicorn has yet to win out.)
It often isn't until we've established tighter patterns and principles that we can take a real stab at tightening up the interaction metaphor. This can be the most exciting moment in developing the system, because if the shoe fits, we can begin to pare away and mold the system to the metaphor (and vice versa). We know the metaphor works if it describes the ideal organizational patterns of an interactive system, the principles of its behavior in use over time, and the emotional impact created via repeat utilization of its features and functionality. Ideally, it must also remain flexible, and be able to scale at least a few orders of magnitude without falling apart. If it does, we know we need to tinker with it some more.
If a metaphor doesn't work—and often it takes a few attempts—we discard the metaphor and try another. If we do, we run with it as far as we can!
Uh… This Seems LIke a Lot of Work…
Not every system requires an amazing parti or a unifying interaction metaphor—it is a construct that helps us consider how the system lives and breathes as a whole. To quote Matthew Frederick from 101 Things I Learned in Architecture School:
Some will argue that an ideal parti is wholly inclusive—that it informs every aspect of a building from its overall configuration and structural system to the shape of the doorknobs. Others believe that a perfect parti is neither attainable nor desirable.
From my perspective, I find such metaphors attainable and desirable. They are foundational thinking, and useful when describing to clients what elements should and should not be disrupted. They are also at parity with other disciplines of design.
Those of us with a deep background in print and communication design have grown quite familiar with this process of establishing the core ideas that inform this arrangement of content, and in some cases, the implied flow from place to place.
Over the past decade, we've seen dramatic changes occur in the graphic design world—much of it influenced by user experience design and industrial design practices crashing straight into the marketing and advertising bus.
In this shift, visual designers have become more aware of how they aren't just crafting artifacts to communicate aspects of big conceptual ideas and themes, but that they are also shaping people's behavior as they interact with those artifacts. In a sense, we're all becoming a bit more like interaction designers, whether we desire the change or not.
Product designers have understood this for quite some time, as they are intimately aware of the potential affordances any real, designed object or series of objects would include. But interaction designers in particular have been exploring the space between those designed objects and the potential affordances they may offer in the physical world. No "big idea," brand pillars, design templates, or snazzy campaign theme can easily determine the potentially limitless range of uses that such affordances may offer. In essence, interaction designers are creating the glue that connects priorities of use with capabilities that may have never existed before, in quite the same combination. In the world of interaction design, "big ideas" can often end up being themes that support and enhance the metaphor of the system.
Because of this, learning to consider how a sheaf of ideas coheres into a interaction model—instead of creating point solutions for client problems that are impacted by interactive systems—is a worthy skill to gain. It will be hard fought to learn, but immensely valuable in the years to come, no matter what kind of design you practice.
And the metaphor that serves as the foundation beneath that interaction model—that will help to govern the emotional qualities that your users derive from that system's use. In those moments, everyday actions become poetry.