Design and long-term thinking

In my work, I’ve been thinking a great deal about time lately. Perhaps this is a function of my age or perhaps it’s just in response to the Silicon Valley saying: “Move fast and break things” which is adorable until you start to hear about how that mantra has indeed broken things. A soon to be renamed Facebook is merely the most recent company facing backlash over this.

In the rush to scale we overlook how the things we build are negatively affecting people.

In the cramped nature of the sprint we’ve compartmentalized our work, forcing delivery at a pace that too often moves at the expense of peoples needs.

In our desire to disrupt, we rush to market, often without considering historically excluded people.

So time is really the issue here. How we think about time and the nature of what we build is very much top of mind, and not just with me. Jorge Arrango is thinking a great deal about this and talked to Cyd Harrel concerning this topic specifically as it relates to her work in the civic space…which has very different time dependencies than necessarily commerce.

But way before my contemporaries started thinking about time, Stewart Brand conceived the Pace Layering diagram (below left) in 1999 (with roots prior in his book, How Buildings Learn.) He wanted to describe how human civilization works in terms of time which was leading to his work for the Long Now Foundation. Fast forward to 2016 and Stefanie DiRusso created a Typology of Design Thinking (below right )which I’ve adapted and have been using for several years to describe a combination of design complexity and maturity.

In order to better understand time as it relates to design, I’m going to try and meaningfully combine the two diagrams.

I’ve been thinking about these models and it struck me that if I formatted them consistently. They could be presented alongside each other. Brand points out that software developers immediately recognized the utility of the original to consider software development in terms of complexity. So I’m merely borrowing these constructs to illustrate how time and design work.

In a presentation with Paul Saffo about the original Pace Layers diagram, Brand responds to a question about logarithmic vs exponential time scales, he said, “This is a data free diagram.” which likely leaves it open to interpretation so I thought I’d provide some time frames (in orange) as a suggested addition. The time frames are also another way to make the diagrams consistent as time was not a component of the Typology of Design Thinking diagram. Also, as mentioned several times in the presentation by Brand and Saffo, the liminal spaces between layers are extremely interesting and are often places where the interaction of the layers can foment change. For this reason, I used dotted lines, rather than solid between the layers.

LAYERS

Fashion. I’ve left Fashion to the original squiggly line. When Brand and Brian Eno first developed the model, Eno argued for Fashion rather than Art as the right label, and there was apparently a lot of back forth about how squiggly the line should be…and who am I to argue with that. :) Also, I’m less concerned with fashion as a design activity.

Let’s say that timeframes for Commerce tend to happen quarterly for publicly traded companies. It’s as useful as any way to start assigning this kind of detail to the diagram. However, this measure leaves out more instances that also exist. Startups think about time to market. They think about burn rates, and later, scale. These constructs are indeed a problem with how some of our products have failed us so perhaps I’ll draw up another variation of the diagram. Mr. Brand stayed away from specifics (his line about the diagram being data free) possibly for this reason.

I chose one year as a time period for Infrastructure. I live in the San Francisco Bay area and the Eastern span of the Bay Bridge was replaced starting in 2002 to 2013. The Salesforce tower in San Francisco took 5 years to build and now dominates the skyline. So it feels like infrastructure exists in yearly increments.

Governance I listed as multiple years. I think of it in terms of the election cycle so one could think in terms of 2–4 year increments. Different countries have different spans of time for their election cycles…if they have elections. Brand is a believer that cities actually hold some power in this capacity and they touch on the notion that a longer-term elected official, can actually get a great deal done. Again, the times can be adjusted for different countries and different perspectives.

Culture is a challenging layer, and I put a longer time frame for it because changing culture can seem quick to us, as when Hamilton the musical appeared. When this happened, it began to change our culture. It took a great deal longer to write, and get produced that it did to begin to change the culture though. The diagram label here could just as easily have been a generation. Technology is changing culture fairly quickly like with the advent of the cell phone and later the iPhone. These are good questions to ponder and to apply to the diagram in order to help it convey different aspects of how we work.

Lastly, I gave Nature a 100-year label. But as Paul Saffo points out, in the Long Now discussion with Brand, a fault line, like the San Andreas Fault, can move 2 inches a year or have cause a devastating earthquake that changes everything in a minute. The climate crises may have started with the beginning of the Industrial Revolution or the advent of cars and big agriculture. Today the effects of the climate crisis are seemingly more noticeable to everyone with the severity of storms in the last several years, the flooding in Germany over the this past summer, and the ongoing drought in California. I’m open to different time representations.

So here is my combination of Stefanie and Stewarts models.

It might be useful for you to step away from this article to review the original if you wish, with Stefanie’s PHD Thesis work. I inverted the orientation of her labels to align the work that designers might do with the time scales in the Pace Layer model. In her original, the more complex things that people would design are at the top, this one has them at the bottom. I also point to different complexity labels than she used in her version.

At the Artifact level are the objects of design. A logo, a shirt, a chair and even a website exist at this level. I have those listed with an S1, S2, etc. to indicate a sprint — a two week increment of work, common in digital product development. Physical products have longer development times than digital ones. One could create a logo in a day. Again, the variation in time scales are to be expected and I would want folks to play with the variations as they make sense to you and your organization.

Thinking about an Experience in terms of human-centered design or user experience happen in slightly longer time scales…or at least they should. When user experience gets lumped together with user interface, the two could be confused as only responsible for delivering at the artifact level. “Oh you’re the UI/UX Designer? Great. When are those wireframes going to be ready, the engineering team needs them.” This labeling of UX alongside UI is wrong.

Alternatively, the true nature of what a UX team should be doing is known, but the work gets slotted into sprints. In this way there may not be time to do the work properly. I’m not against agile or sprints per se, but UX and QA usually get truncated in the process, reducing their time to explore and provide a solid experience.

Systems is a word I use to correspond to infrastructure (from Pace Layers) as they take more time to build or revise. When I started at McKinsey in 2015, onboarding with a new colleague my first day. When we opened our laptops, I looked around for Outlook. The HR person said, “You’ll find Lotus Notes in the Start menu…” — and I looked at my colleague wide-eyed. Lotus Notes still existed? McKinsey was using it? I asked the HR person. They said, “Yes, but we began beta testing Outlook recently.” It took another 3 years for McKinsey to make the switch.

There’s a lot of reasons for this, but suffice it to say, it takes time to install a system like Lotus, train a workforce, iron out the kinks, and ensure it works. And although seemingly arcane to me, it worked fine. It can take years to dismantle an existing tool and retrain everyone to use a something different. It can take as long to design a new system from scratch.

Policy in this diagram equates to the Governance layer in Pace Layers. I assigned the same time scale to this diagram and I strongly believe that although lawmaking and policy writing ins’t traditionally a designers job, the people doing this job are (ideally) intentionally trying to problem solve. I think that everything is designed. Lawmakers sit down and write legislation that they argue over. They add tangential or disparate material or remove items until the end result is often an unrecognizable aggregation. In some cases the legislation solves problems for constituents, but often not. These activities are designed in as much as the lawmakers are trying to problem solve.

Social infrastructure or how we deal with each other is akin to culture in Brands’ diagram. And although this has a longer time frame, it is indeed designed. Often by omission or over such a long time and by so many people that it’s unrecognizable as design. But again, to Paul Saffo’s point, sometimes layers move faster than the diagram portrays and the intersection of layers are points of change.

Lastly I have Environment. Humans have had a crazy and deleterious effect on our environment. In this other article I wrote, I point out how the colonizers of the land we now call the Americas nearly completely destroyed every wolf and buffalo they could get their hands on. The reintroduction of these species back into the environment of Yellowstone, a National Park in the US, has had an amazingly positive effect. The same is true globally. People, often referred to as the anthropocene, have affected our global environment drastically in ways too numerous to mention here.

Now, we can begin to map individual instances.

MAPPING DESIGNED INSTANCES OVER A LONGER TERM

A (in red) on the left of the diagram could be anything, but I chose Facebook’s NewsFeed as an example. In 2006, Facebook introduced the NewsFeed as a feature of its service. This was originally a scrolling list of content from your friends. Later they added traditional news to the feed, a ‘like’ button (thumbs up icon) that has become iconic of the company, and of course, they introduced advertising as well.

What started as a seemingly innocuous feature at the artifact level became a social change agent.

The feed encourages engagement, which is paramount at Facebook. Engagement means time in the app and views/clicks on advertisements which is literally how Facebook makes money. Advertisers pay Facebook to appear to specific people (through demographics) in their feeds. Forgive me for oversimplifying this, plenty of others have written about this model.

The ‘like’ button helps the algorithm behind the items that appear in the feed and encourage simple and additional engagement. The user experience scrolls content and you don’t need to click away from the feed to quickly “like” an item. Super simple.

Because the feed was largely based on what your friends were doing or saying, you ended up in what has become commonly known as a social bubble. Your friends post what they do or like and those things are of interest to you…because they’re your friends. The algorithm examines this info and provides articles that you and your friends would appreciate and this forms a feedback loop where you rarely look outside of this “bubble” of content. The bubble shields you from differing opinions.

Here are some articles on the recent Facebook issues from the Washington Post, Sky News, and the New York Times.

In short, Facebook has had numerous people from within the company come forward to indicate, with substantial documents, that they knew a great deal about the negative aspects of the bubble including the spread of lies on the platform. All this started with an artifact, the NewsFeed, but it permeates the layers in the diagram above and lands in the social infrastructure layer because the lies spread through the platform to billions of people are in many ways, killing us.

B (in green above) is a positive example. “At Long Now, we often like to tell the story — or perhaps better said, legend — of the oak beams at New College in Oxford. First told to Stewart Brand by anthropologist Gregory Bateson, this short and simple story epitomizes the tremendous value we can reap from some long-term thinking.”

The story is that the huge oak beams at the school were starting to reach the end of their lifespan. Someone suggested there might be oaks on college property and when they went to ask the gardeners about it, the gardeners said, “We were wondering when you’d come ask us as we’d planted oaks just for this reason on the grounds to replace the ones you need.”

Here the artifact is a set of trees. Planting them enriched the environment and added to the aesthetic of the landscape. It served a genuine and constructive purpose but took time. I placed the end point in the environment section, but the story is about a systemic view with a circular application. By circular I mean that, likely, trees were cut down to create the original beams. A grove was held aside to replace the first set of old beams, and ostensibly, new ones would be planted for the next time as well.

This is the kind of long-term thinking we need. Moving fast and breaking things is indeed that. But breaking things in the case of the Facebook example is detrimental to the health of governments, our response to the pandemic, and how we treat each other as humans. How can we apply this positive instance to a digital example?

We can start by taking more time to examine and design the things we’re building and we can prioritize ethics over profit.

CONCLUSION

I’ve learned several things from this exercise.

1. By putting real times in the diagram, we can think about the items more tangibly. As I’ve stated numerous times, the numbers I used are not a one size fits all, but if you contact me, I’ll send you a copy of the diagram and you can substitute your own numbers and make other changes.
2. Thinking about time for the work we do as designers helps us to better understand what it means to design and what changes we can affect and when. This was the reason for combining Pace Layers and the Typology of Design Thinking models to begin with.
3. In our effort to scale these wonderful digital tools, we’ve left huge gaps — and the tools are working poorly for a lot of people. In the case of Facebook for instance, the speed of commerce has disrupted governance. Arguably worldwide in different ways. By moving at the speed of commerce, we miss opportunities to achieve better user experiences through a more thorough examination of the interactions we’re trying to solve for.

Thanks for taking the time to read this.