by Kate O'Neill
Responsive Design and Integrated Human Experience Design
There’s been a movement within web design over the last few years advocating for “responsive design”: designing content that incorporates adaptive styling that adjusts appropriately when viewed on a range of screen sizes. Instead of having a “mobile version” of your website, for example, your content is simply your content, and it is prepared in such a way that the same content will work equally well on a 27-inch monitor and a 4.7-inch smartphone screen.
While that’s not the same premise as designing with an awareness of the physical context someone is in—or agnosticism about their physical context, depending on how you look at it—there’s certainly a related idea going on there. In the case of responsive design, the presentation of a digital experience allows for the possibility of multiple device contexts.
In Integrated Human Experience Design, the presentation of any experience, online or offline, allows for the possibility of multiple digital or physical contexts. In other words, if I’m designing a shopping experience that is primarily focused on the physical retail space, I must also be mindful of the digital interaction the customer will have with the store, and allow for nuanced opportunities to enrich the shopping experience through the integration of those layers.
Data Determines Success; You Determine the Data Model
There’s always some benchmark for whether or not a project worked, or how well it worked, and at some point in a for-profit corporation that benchmark, ultimately, is bound to be profit. But there are meaningful measures of preliminary success well before interactions turn to profit, and usually they follow some sort of “funnel” model. Typically, the data points the company tracks relative to that funnel aren’t customer-centric but rather company-centric, like the old standard: “awareness > consideration > preference > purchase.” In almost every environment, some metric will be the benchmark of success in every meeting and every debrief of that project.
The data we collect about that project is, then, how we evaluate the project’s success.
Here’s the thing, though: You are (or someone you work with is) determining or helping to determine that data model. Which means there’s a chance to define the success in terms of how the customer experiences it. Some companies are doing this, and taking a truly customer-led approach to defining the success of their brands.
Did the customer find what they were looking for? How do you know? Did the customer complete the task they came in to complete?
As I often say: Analytics are people. And relevance, in terms of offering targeted messages and experiences, is a form of showing respect for your customer’s time and interests. So is discretion regarding their privacy. All of these principles are really about data practices.
We all know that companies are going to collect customer data, so I want to equip them with frameworks for doing it ethically, mindfully, meaningfully. We all know that companies are going use customer data to sell products and services; I want to help them do it in a way that centers humanity in the perspective.
It’s important that you then have the mindset as you approach the data model that the data needs to be rich and relevant, and it needs to be humanistic and able to encompass the richness and rigor of the human experience.
CHAPTER FOUR
The Humanity in the Data
If you take nothing else away from this book, what I most want you to remember is that whenever you see website analytics, or customer data, or big data about health trends, or any other data that has to do with something human beings have bought, done, or said, you should stop and think about the humanity represented in that data.
That’s it. That’s my big goal.
In the many years I’ve worked with companies on their marketing strategies and digital analytics and user experience and so on, I’ve noticed how common it is for marketers, executives, designers, and even strategists to get caught up in the abstractions of the data, and forget that they’re talking about real people. That conversion optimization is about getting more people to buy something—which is great, if what you’re enticing them to buy has value to them. If you can find the people for whom the thing is valuable, you’re doing them a service. If not, you’re making the world a worse place. Don’t make the world a worse place.
I have said, and often say, that analytics are people. And of course I know that’s an oversimplification. There are plenty of analytics about server logs, manufacturing equipment, industrial machinery, and so on that appear to have nothing to do with people. But then again, don’t they? Why else would we care about the efficiency of those servers, the output of that equipment, and the safety of that machinery if not for people? The key idea here is that what we measure, we typically measure because it makes our human lives better. And it’s good to remember that.
We often hear experts advocating for empathy in the design process, but rarely do we hear about it in analysis. Marketing cycles of campaign and channel management bring with them a whole lot of access to consumer data, but unless the marketer is willing to see the data as a proxy for actual people, there’s no empathy. Really, the more data a marketer has, the more empathy they should be able to feel . . . if they’re correctly associating the data with the humanity on the other side.
The Data Layers that Connect the Physical and the Digital
Increasingly, our surroundings are created by actors we don’t know are there. We can’t see them, we probably don’t understand them, and we have no control over them anyway. So when it comes to the convergence of the physical and digital worlds, it’s even more important to remember to look for the human actor because they’re the ones who primarily create the data layer that connects these worlds.
It’s always been worth considering context in the design of experience. Content and interactions resonate best when they seem to flow naturally, which means considering where a user (or shopper or patient or person in some other role) might start, how they might continue, and what they will consider a successful finish.
It’s even more important to consider context when it involves the physical reality of a person’s surroundings, their device and bandwidth limitations, their environmental distractions, and more.
With or without connected devices, relating lived human experience to a virtual abstraction of that experience can lead the way to the discovery of meaningful patterns within human-generated data.
The Human Component of the Internet of Things
The buzz about the “Internet of Things” has been going on for a while. Fundamentally, it’s about connecting devices to the internet and web, and to each other. It’s any item that could/can use an internet connection.
The big deal is about the data it can/could send.
These devices potentially have a wide range of uses: for automation of various personal or business processes; for just-in-time reminders about purchasing supplies; for proactive maintenance; for intelligence insights into usage or related data with other devices.
Sixteen billion connected devices are forecasted to join the Internet of Things by the end of 2021, and by 2018 there will be more connected things than mobile devices according to Ericsson13. According to McKinsey, six types of applications are emerging in two broad categories: information and analysis and automation and control.14
Information and analysis includes tracking behavior, enhanced situational awareness, and sensor-driven decision analytics. Automation and control includes process optimization, optimized resource consumption, and complex autonomous systems.
Sometimes these devices are also called “smart” devices, but the only thing that makes many of them smart is that they have some data-enabled threshold for alerts, for modifying settings, or for responding to external stimuli in a specialized, programmatic way. For example, a normal, non-internet-connected thermostat can react to external stimuli like increasing room temperatures by also increasing the fan speed of the air-conditioning and cooling the room until i
t reaches a set temperature. A “smart” or internet-connected thermostat, though, will have that same ability but also may have the ability to: check local forecasts to see if the outside temperature is increasing or decreasing; determine from your settings how long it will be before you’re home from work; or vary the energy expended to cool the room accordingly.
In other words, they’re still “smart” in only a very limited way. They’re more cleverly useful than they are smart.
(By the way we’re awfully quick to call connected devices “smart,” while at the same time our culture and media made intellectualism a challenging characteristic to display. What is the benefit to claiming “smart”-ness for our devices and not for ourselves?)
Here’s the thing, though.
The term Internet of Things is at least partially misleading, since a great deal of the “things” only have relevance to human interaction and experience. Our movements and interactions generate data points, in a vast and expanding cosmos of data points, which only truly relate to each other in how they characterize and give dimension to the human experience. Otherwise, who cares? Who wants to know anything about what a smart refrigerator says about cooling patterns or its contents unless that information, in concert with a wearable fitness band, perhaps, paints a picture of the human life it services, and what that person chooses for nourishment?
In other words, it’s still what it all means that matters. No raw data on its own is particularly interesting until a pattern emerges. Patterns indicate choices, actions, trends—human actors behaving in whatever way they behave. And we, the analysts, the sociologists, the marketers, the observers, are here to examine the data for those patterns, to ask the data what it will reveal about our fellow humans and their wants and needs and tendencies and fears and annoyances and habits.
Of course meaning is subjective, as it is in every case. And here is the difference between significance in a statistical sense and significance in a storytelling sense. Perhaps a data pattern emerges that points to a clear, unambiguous trend in human behavior. We can quantitatively observe this pattern, and know that this trend exists. But when we’re exploring the qualitative side of the story, trying to deduce meaning from the trend, we potentially run into trouble. Because our own biases are likely to get involved in the narrative. Our own behaviors weave into the process.
Kids and data
There’s also another consideration: kids.
In fact, parents often post information about their kids even before they’re born, creating a data trail around them from moment one.
Internet-connected baby monitors, tracking systems at schools, and all sorts of other data-intensive devices and systems are creating the data-heavy framework kids are growing up into.
It’s understandably nightmarish to think of privacy and security violations where children’s devices are concerned. A couple discovered that a stranger had hacked into their baby monitor and was spying in on their three-year-old toddler, and even saying disturbing things to the child through the speaker.15 The hacker used the night-vision lens to follow their movements in the room.
There are starting to be a lot of “smart” toys, and that’s attaching data to kids’ usage patterns long before they’re old enough to post ill-considered selfies.
For businesses that market to parents and sell children’s toys, clothing, and other stuff, targeting is tempting. But we need to tread extra carefully around this.
The Data Trail: Everything Has Data History
I’m sitting in a hotel lobby working on this book, and conscious of the data layer, the digital experience layer all around me. People sitting in the lobby are on their phones and laptops, browsing and interacting on social media, checking in on Foursquare, checking flight times, ordering food to be delivered, making dinner reservations, buying odds and ends to be delivered to them back home, playing online games, texting with friends.
And all of it generates a data trail. All of it is trackable somewhere at some level, and much of it is traceable to this location. A good deal of it is happening on this hotel’s Wi-Fi network.
Someone posts a photo to Instagram, and its location is tagged as the hotel. And maybe it has the hashtag “#nashville.” Or maybe “#musiccity.” All of the photos being posted today with those hashtags are connected at the joint of their metadata, and in a way, all of the people who posted and who will like those photos are therefore also connected by that metadata.
That’s the most interesting part of it, isn’t it? How we get connected to each other by the data we create. How we create an exoskeleton of bits around us through the location tags and geo-mapped transactions we engage in. How we live our lives a little more split every day between two worlds: the purely physical one around us, and the virtual/digital one that exists on screens and in the ether. We tap into our virtual world amid these fleeting moments of connectedness.
None of this is news to you, of course. You know we have FOMO and digital detoxes. You know the distraction of inboxes and notifications and likes and direct messages. You know the struggle to be present with your family and friends while your phone is lighting up with alerts.
It’s not just the tension to live in these two worlds that makes this interesting, though; it’s that we are the layer that connects the two worlds. There is no point to big data without humanity and human interaction and human experience. Sure, machines and systems produce mountains of data in their autonomous functioning, but their functioning is usually somehow in service of a human outcome. Manufacturing machines track their own production, but the point of their production is human consumption, by and large.
***
Everything you see in the physical world surrounding you has a data history. I see coffee on shelves, and its production, from roasting to packaging, has had digital tracking components. It has a data trail.
Everything you see in the physical world surrounding you has a data history.
What makes this data trail interesting is not the coffee. It’s the fact that the coffee is intended for human consumption. Everything is tracked not for the sake of tracking coffee, not for the love of data, but so that it is profitable for humans, purchasable by humans, and enjoyable for humans.
But even more so, we ourselves have a personal data trail, which we leave through our interactions with our phones, with cash registers, with security cameras, with sensors, with our cars and transit, and more. Or, as James Bamford put it in Wired when writing in 2012 about the National Security Agency’s data center in Bluffdale, Utah, “parking receipts, travel itineraries, bookstore purchases, and other digital ‘pocket litter16’.”
There are, of course, huge privacy implications of that trail.
***
Increasingly, we spend our waking lives tethered to the digital devices that free us from the mundanity of being online only while seated at a desk. (And for that matter, many of us are connected while we sleep.) These devices by design collect and transmit data about us, logging our surroundings, our circumstances, and our behavior at various levels of abstraction: our kinesthetic data (think: apps that use an accelerometer), our biometric data (think: apps that track how many steps we take or how many flights of stairs we climb), our auditory environment (think: apps that collect a sample of music and identify a song), the frequency and timing of our interactions with the digitally connected world, and so on.
These devices are also connected to each other. And “Things” (as in “Internet of”) that don’t even register to us as “devices,” such as cars and refrigerators, are increasingly tracking their own data around us; and that data potentially meshes with the data our carried devices are tracking. There’s a staggering amount of it (data, that is), transmitting rapidly, silently, and invisibly, structurally describing our environment so that it can be parsed and understood, and opportunistically reshaped by analysts, technologists, and designers.
And that’s only the passive stuff—the data tracked as a byproduct of using
your iPhone or wearing your FitBit. We haven’t even touched on the oceans of data we actively create with our Facebook pictures of Elf on the Shelf, our Spotify playlists, our inspirational quote tweets, our endless retro-filtered latte foam pictures on Instagram.
According to research compiled by the data analytics company Domo, in every minute of the day, there are four million Google searches, 204 million emails, over 2,460,000 Facebook posts, and over 277,000 tweets. Then there’s the seventy-two hours of video uploaded to YouTube each minute, not to mention activity on Instagram, Tumblr, Flickr, Foursquare, and all the time and money spent shopping online. In fact those numbers are more than two years old, so you can bet they’re now even higher.17
To these we add hashtags, links to friends, comments, hearts, likes, and favorites; and what potentially emerges, if we are to assemble the data just so, is a connect-the-dots picture of our projected selves. The Self on the Shelf, as it were.
Of course this raises questions about privacy. How do we ensure that our most personal and vulnerable information is kept from systems that could abuse it? It raises questions about security. What decisions do we make about the devices we use and the systems with which we choose to interact based on how safe we believe them to be? What determines private? What determines safe? What determines a reasonable trade-off for private and safe? Is it convenience? Is it cost? Do we make these decisions intentionally or without consideration?