“If it disagrees with experiment, it is wrong.” —Richard Feynman

More than a year ago, we set out to create a new approach to digital work that would solve as many of the pain points as we could for both our clients and ourselves.

I recently came across this short clip from a 1964 lecture in which Nobel-winning physicist Richard Feynman distills the essence of the scientific method, and was immediately struck by how closely it describes the way we work.

We didn’t set out to replicate the scientific method, and there’s certainly quite a bit of art to our science, but the approach is pleasingly similar: make an educated guess, test it relentlessly, and adapt as needed until you arrive at the right solution.

Effective solutions on the web may be moving targets, but so long as we’re paying attention we can move with them.

“I get an idea from my head into your head by vibrating the air through flapping the meat in my lips. Very clever hack.” — Jeremy Keith

In the world of cinema, there’s a concept called Auteur Theory which postulates that a director’s creative influence can be so strong that it pervades all aspects of what winds up in the final theatre experience, despite the highly collaborative nature of the medium.

It’s an interesting theory and does much to explain how the collected works of, say, Alfred Hitchcock can all seem so… Hitchcockian. Many directors, working through others, have left behind creative fingerprints every bit as unique and identifiable as Van Gogh alone at his easel.

For this topic, however, a key phrase there is “left behind” – Citizen Kane, an oft-cited work of auteur genius, hasn’t changed in the 70+ years since the final cut was sent to print. For cinema, the creative evolution of the project comes to a close long before it’s ever shared with the general public.

Web projects, on the other hand, require a more organic and ongoing creative process. The web itself, by its nature, is never really “done”. Unlike film, a website’s creative process is only getting started when it’s shared with the general public. A website remains in need of a guiding hand long after it has launched and those that had originally conceived it have moved on to other projects.

Unfortunately, many web projects today are run in such an “auteur” fashion: one person assumes intellectual ownership of the project and executes their vision by delegating out discrete parts to their collaborators. This scenario requires that person to shepherd their vision through the contributions of others, effectively asking them to “make it as I see it”. This situation can be compounded by a siloed design process, forcing the “vision-holder” baton to be passed at each phase checkpoint.

Hopefully the collaborators can get a clear idea of what’s in the visonary’s head most of the time, but this approach leaves things highly open to interpretation throughout the creation process and a tremendous amount of intellectual value leaves with the visionary when they move on to another project. The auteur system simply fails to work as soon as the auteur has left the picture.

That’s why we at The Good are so big on decentralizing the intellectual ownership of a project’s essence and direction from day one. We replace a project’s “visionary” with a Project Vision – a document for all participants (current and future) to collaborate around. “Make what I see” simply doesn’t cut it when building goals and not simply features.

Beyond the Project Vision, we are careful to keep all our design documentation focused on defining the understanding of the problem space rather than the attributes of the solution we imagine addressing it. This provides everyone involved the proper context to contribute in their own expertise without simply taking stabs at faithfully producing their interpretation of someone else’s vision.

But even beyond that, we place a premium on taking every opportunity to get our thoughts out of our own heads and into a shared space where they can evolve. We’re not in the insight-hoarding business — we’re in the value-providing one. The shortest path between the two is simple: share your ideas. The vision’s only as good as where it takes you.

“A business absolutely devoted to service will have only one worry about profits. They will be embarrassingly large.” –Henry Ford

Just the other day, my family’s go-to digital camera finally gave up the ghost. I didn’t want to drive to a big box retailer and hope my luck outmatched my patience, so I hired a personal assistant help me pick out a new one instead.

I asked them to show me some options in my price range, then had them point out some models that other camera buyers recommended. I narrowed it down to two candidates and asked my assistant to tell me what owners of each model had to say about them, then picked one and had my assistant ship it to me. A couple days later we were a happy, photo-shooting family again.

Of course, I didn’t have a human assistant doing these things for me – I’d hired a website, instead. In this case, Amazon.com was a great help in what could have otherwise been a drawn-out and dissatisfying experience.

All of us are faced with situations throughout the day that we look to the internet to resolve for us. Want to know if that breakfast place is open yet? Hire their website to find out. Need to see what the score of the big game is? Hire ESPN.com to tell you in realtime. In the mood for some amusement? Hire FunnyOrDie.com to show you some entertaining videos.

If you think of every online interaction you have as a “hey assistant – do something for me” type situation you will begin to see that that dynamic constitutes roughly 100% of your internet experiences – you rarely go to websites for no reason. In fact, I’m going to go ahead and say you NEVER go to websites for no reason. Let’s call that reason your “intent”.

As web designers, we would do good to remember that discovering and supporting the users’ set of intents is the first order of business – if we can’t deliver on that, we’re not really going to get much of anywhere. The user recognized they had a need, thought you could serve it and pulled up your site: you can either be helpful… or be irrelevant.

Unlike marketing efforts that impose themselves on our attention space like billboards or TV commercials or the (regrettable) use of human beings as street advertisements, websites must quite literally be requested into existence – they’re just ones and zeroes sitting on a server somewhere, waiting to be called up and displayed. And it is that act of requesting – of a user recognizing a need and acting on it by calling up your site – that constitutes the essence of an online interaction. Failing to support that intention undermines all other brand strategies you could hope to achieve.

In the world of advertising, first it was “push” – marketing by interruption. Then the stranglehold on our attention dispersed and it became “pull” – marketing by attraction. Now with more and more opportunities for engagement moving into the interactive sphere, the chief marketing strategy is “serve” – marketing by being helpful.

For websites, you’re only lucky enough to exist if the audience intends to make use of you. It’s up to us as designers to understand & serve that intent as best we can. We should happily accept our contribution’s role as the “assistant” – there are a lot of openings in that sector these days.

“The greatest compliment that was ever paid me was when one asked me what I thought, and attended to my answer.”
— Henry David Thoreau

Getting user feedback is hard.

It’s so hard that companies pay war chests of hard-earned revenue to consulting agencies who wheedle it from their audience. They drag Average Joes into two-way mirror testing chambers, invade their most private social networks and inject popup surveys between them and their content at every permissible opportunity, all in the name of finding out what is meaningful to them.

So, you’d think if there were an abundance of voluntarily-provided feedback, they’d treat it like some kind of godsend and react to it with equal parts compassion and excitement. Instead, we get the modern FAQ page.

I’m just going to come right out and say it – 99% of FAQ pages are built on two kinds of bullshit: lies or laziness.

Let’s set aside the mind-bogglingly inane practice of inventing your own “frequently-asked” questions (the “lies” part) and assume your clientele truly is persistently barraging you with the same set of inquiries, day in and day out.

If people are constantly asking you the same questions, it means two things:

1. These questions are important to your audience, and…
2. Your website is doing a crap job in answering them

Your website’s users don’t want to contact you to get the information they need — that’s why they’re on your website to begin with. For them, reaching out is a last resort. And for each that does, many others won’t even bother.

Of course, we can’t predict every single aspect of an interaction ahead of time and users are bound to surprise us with things we hadn’t planned for. It’s what you do with your post-launch findings that matters, and putting them to applicable use is a golden opportunity for improvement. Shelving them away in a dark corner of your website is not a particularly great way to seize the opportunity, to say the least.

Like your customers’ resistance to reach out, it should also be a last resort for you to create an FAQ page. In most cases it’s a cop-out, a white flag of surrender. Most FAQ pages are basically saying “We give up. There’s no way to design this site to accommodate user needs X, Y and Z — let’s just toss ‘em in a catch-all bin.”

So what can you do instead? Treat user feedback as what it is: a valuable list of potential improvements you need to investigate and, usually, adapt your site around.

Sick of people asking you how much your product costs? Find a way for the website to better answer that question ahead of time. Getting pestered with tedious requests for info already available on the site? Could be time to find ways to surface it using a more resonant language or content structure. Annoyed with all of the questions about how to recover lost passwords? Sounds like the current sign-in component isn’t doing its job.

Your website exists to serve the overlapping needs of the business and its customers. It exists in a continually changing social environment, and that change needs to be met with equally continuous internal calibration. Let’s take every opportunity we can to genuinely attend to the feedback we value, stop guessing at what people might want to know, and start paying attention instead.

“We don’t see things as they are, we see them as we are.” – Anais Nin

I have always thought baby mobiles are designed backwards. Let’s say, for example, we have a very cute and pleasant one with colorful wooden cutouts of bunnies. To the adult who installed it, this looks perfect as it hangs over the crib:

but to the baby (the party this product is supposed to be designed for) it projects a much less interesting visual:

There seems to be a tension here: its purpose is to entertain babies, but it’s positioned to appeal to adults. While having the bunnies face the cognizant, wallet-wielding side of the equation may be the more marketable decision, the baby’s experience is no doubt the poorer for it. In fact, in this particular case their enjoyment of looking at the bunnies is at the direct expense of their child’s ability to do so. The parent feels they’re providing something enjoyable to the baby, but objective observation indicates otherwise.

This tension is no different in the web world than it is in the product world – clients (those with the money) want to receive a website that they like, assuming that that feeling will be mutual amongst their customers (those experiencing it). If we turn the dial too far towards the client, though, we risk providing a spectacular but irrelevant product – a virtual “yes man” of a website, comforting them but not serving their best interests.

In my mind, this is the essential crux of user-centered design: identifying the tradeoffs of catering to either the client or the user, and presenting suggestions for how both parties may best be served.

When in doubt, it seems like the best option is to side with the people actually using it – not only are they the ones it’s presumably supposed to be designed for, but their perspective is also likely to be the least-represented throughout the design phase. Moreover, by serving the users we in fact serve both parties simultaneously: the website is only valuable to the client if it’s creating a real-world effect in behavior, and that can’t happen if it doesn’t resonate with its audience.

You would be hard-pressed to find a parent who would sign off on an instruction like “make this baby toy less interesting to my daughter and more interesting to me”, and yet examples abound, particularly online. The hope of user-centered design is to remind us of our higher ambitions, and remember which way the bunnies should be facing.

Most clients, especially those working with traditional design & marketing agencies, are accustomed to a “waterfall” process, in which their project is handed to a large team and divided into a series of discrete phases such as discovery, design, development, testing, launch, and (with any luck) maintenance.

The waterfall process offers several perceived advantages for both agency and client, however many times those benefits are outweighed by the inherent inflexibility of the system. For instance, once a client has provided sign off on a design layout, any change no matter how small or logical typically triggers a change order and schedule adjustment. This is just one example of the back and forth that naturally occurs with any digital project where the waterfall method is not well poised to respond naturally.

Clients and other project stakeholders may not be able to fully understand how a certain feature will work — or even if that feature is necessary — simply by looking at a wireframe or a design comp. Features can be added in a few minutes of design that can result in days or weeks worth of programming. A seemingly simple content change can have a ripple effect throughout the structure of a site. All of these result in costly changes to the project and timeline.

Just because things have been done a certain way for years doesn’t mean they can’t change, and change quickly.

We believe in a “lean” process in which we work to eliminate wasted time and effort by delivering work for review as quickly as possible, and gathering feedback to incorporate new knowledge into the next iteration of the project.

The leaner we are, the easier it is for us to respond to change. Using smaller, more focused project teams gives us something special: agility. The ability to be nimble and respond quickly and appropriately to change is one thing that small teams have by default that large teams don’t. A change that might take a larger team in a large agency a week to implement might only take a day for a smaller, leaner team to deliver — especially if that team is planning to work that way in the first place.

Iterate quickly. Iterate often. Unlike the waterfall process that requires agencies to “get it right the first time”, an iterative process allows us to get a working piece of functionality in front of actual users as quickly as possible, even if it’s not perfect, and adjust from there. We can quickly gather feedback, determine if we need to make changes to the design or functionality, and create solutions to address any concerns for the next iteration. A great example of a successful iterative process at work is this Nordstrom Innovation Lab case study.

Though this is a challenge to how most agencies approach their work, we believe it is one worth taking on. Just because things have been done a certain way for years doesn’t mean they can’t change, and change quickly. We believe that starting from a goal and working toward a feature, expecting and embracing change along the way, will lead to better outcomes for our clients and ourselves.

“Writing about music is like dancing about architecture.” — Elvis Costello

When the art of cinema was young, it was defined in terms of other preexisting art forms – “music of light”, “painting in movement”, “architecture in motion”, and so on. This was reasonable, considering it was in its infancy and its true identity was still emerging. Like any new medium (early television was called “radio with pictures”), the internet has gone through a similar process of self-discovery.

Much like cinema initially referenced the traditional form of theatre before coming into its own, web design has always borrowed heavily from the more established print design world, replicating its values and limitations alike – essentially considering websites to be “documents with clicking”.

However, interactive experiences are terrifically complex and richly nuanced, and crafting the means for such an experience is a tall task. Outside of some documentation remnants of a UX/planning stage, nearly every step of the average site design process (wireframes, roughs, comps, revisions, etc.) concentrates primarily on documenting how the website is presented under a number of different conditions.

Relying on the visual design to do the heavy lifting of the creation process is a problem because most of the non-visual thinking is lost as soon as the layout is handed to a collaborator, leaving them to guess at the intent behind the decisions made. It shows, but often doesn’t tell.

As in the quote at the top, “dancing about architecture” is not the most natural endeavor, and it’s time we questioned whether “compositioning about websites” is either.

Just as cinema went on to attain its unique identity after its basis in replicating theatre, if we peer beyond the graphical aspects of websites that are most readily familiar to us we can see the emerging identity of web design: interaction.

If instead of focusing the project around its layout we all focus on defining and facilitating the behaviors we want that layout to elicit, the project’s conceptual integrity is shared rather than left to interpretation. With a blueprint of behaviors directing not only the visual but also structural and interactive elements, every discipline involved can do their best work without any loss of information along the way, including after launch. This opens up the opportunity to better collaborate across roles, teams and even time.

Websites exist to solve problems – to spread information, to forge relationships, to help people do things they could never have imagined before. As such, the essence of web design lies less in the pixels and content on the screen and more in the real-world human behaviors it facilitates. Shifting our focus from the design of documents to the design of interactions frees us from our commitment to defining a website by its material form and gives us the ability to define a website by what it can and should be: experiences with purpose.

Design the interaction, and the rest will follow. After all, websites don’t even really exist unless they’re used: it isn’t just what it looks like, it’s what it does.

We are rich with theories and hard at work in finding what’s possible in this approach, and we welcome you to join us.

Gut feelings are locked away in a part of the brain that doesn’t have access to language and most attempts to justify or explain them only result in a story spun for ourselves or others to believe. This presents an obvious problem for us as makers when creating a digital project: guidance and feedback can be difficult to provide or receive without resorting to opinion, making it very difficult not only to determine the success of a project, but also to plot a course for success in the first place.

By outlining a series of goals to build on rather than a list of things to build, it’s much easier to determine whether or not an idea will contribute to a project’s success.

How many times have you scheduled a visit to the doctor armed with a well intentioned list of suggestions to solve your problem based on a bit of googling? That’s not much different than hiring an agency to build your website and coming in the door with a list of features; in both cases the expert is being asked to fill a prescription without a full understanding of the problem.

Unfortunately, most projects that begin with this foundation don’t end with a sustainable, measurable outcome that justifies the investment required to create them. After years of experiencing various sides of this arrangement, we’ve shifted our approach from planning the “what” to understanding the “why”, from the features of a project to the goals behind it.

We’ve found that by first outlining a series of goals to build on rather than a list of things to build it is much easier to determine whether or not an idea will contribute to a project’s success. When a feature has to support a business or behavioral goal in order to make the cut, the inevitably varying opinions guiding a project are also forced to align behind a common standard of qualification. This makes actively seeking the simplest and most intuitive solutions to the problems we’re being hired to solve a much clearer process, and though it eventually results in a list of components to build, both we and our client can be confident that we’re on the right track.

Another strength offered by shifting from features to goals is the ability to intelligently correct course after launch. While many projects end with a “post mortem” phase where everyone considers what went well or could have gone better, the last thing we’d call the launch of any digital project would be its death. Unlike a magazine or poster, digital work can be measured and adapted for the better after it’s launched based on real data.

Even after establishing goals and building only the features that support them, the launch of a project is still essentially a best guess at what will succeed. If proper care is taken to set the right goals and build around them, the launch of a project is the first opportunity to take actual usage data and further align the features behind the goals. For instance, if users are frequently searching for a certain type of content, that’s an opportunity to address the site’s content structure. If a certain product or workflow isn’t converting to sales, there’s an opportunity to run some A/B tests to increase conversions.

Figuring any of that out and making the necessary changes requires a solid foundation of clear goals at the outset and follow through after launch. This may seem like common sense, but it’s not so common in our industry and it’s time for that to change.