Designing to preempt a user's every want and need
A few weeks ago I decided to make a reservation at one of Berlin’s finer restaurants. After spending a good amount of time checking out the reviews on Foursquare, I eventually settled on a place called Katz Orange. I hopped onto OpenTable and found a table for two at 8 PM. Score.
In less than a minute I got a confirmation email in Inbox, informing me that the restaurant had received notice of my reservation and was eagerly awaiting my arrival. Shortly after, my phone buzzed and displayed a Google Maps notification, telling me to leave at 7.30 at the latest to be on time for my reservation.
Upon tapping the notification, I was presented with the distance to the restaurant and a short reminder of my reservation. In addition, Maps suggested ordering an Uber for a comfortable trip to the restaurant.
After a short and relaxing ride, the Uber dropped us off right at the restaurant. Perfectly on time. Just as Maps had estimated.
In this scenario, Google was able to correctly anticipate and answer my needs before I even realized I had them. The reservation info was extracted from the email in Inbox, added to my calendar and linked to Maps, to later prompt me to leave on time and suggest taking an Uber. By leveraging the data available across multiple services and devices, Google managed to always stay one step ahead of me and satisfy any of my wants and needs in this situation.
The anecdote illustrates what is likely the next big step for design: Anticipatory Design; Leveraging user data to assess, predict and answer their needs ahead of time. Designing to always be one step ahead.
Not only did Google anticipate my needs in the highlighted scenario, they also never presented me with the choice to be reminded of my reservation. Nor did I request an estimated travel time from my apartment to the restaurant. From the moment I made the reservation to the point that Google prompted me to take an Uber, I was never once asked whether I would like to use any of these services. In fact, I wasn’t even made aware that Google’s integration with these services existed.
And that's okay.
Rather than informing me of all the available options and forcing me to make a decision on whether or not I'd actually like to use these services, Google made this decision for me. And the experience ended up being delightful. Less choice, or no choice in this case, ended up being the better choice.
As we move towards a “future without choice” as Joël van Bodegraven puts it, design decisions become less focused on curating and presenting a selection of options. Rather, the focus is shifted towards preempting the user's every possible need.
This goes against the conventional conception of product design. Traditionally, designers create a product that users interact with, whether it's by clicking on or physically touching the product. The user is presented with a specific amount of information and a range of options and is expected to choose one of these options. In products that are designed to anticipate the user's needs, this decision-making is taken away from the user.
This decrease in decision-making for the user is a deliberate choice and a core characteristic of Anticipatory Design. It’s aimed at lowering the cognitive load of taking almost 35,000 decisions a day. By automating the process, the stress of actually picking from any number of options is alleviated. It’s all about designing systems that intelligently make decisions on behalf of the user.
In the past we’ve seen companies like Netflix and Amazon build systems that recommend new series or movies, or highlight products that we’d probably like to buy based on viewing habits and past purchases. These types of recommendations are only the tip of the iceberg. They still require us to consciously make decisions from this tailored selection of options. The final decision still lies with the user.
In many cases, these platforms complicate decision-making even further. Rather than having to choose from a large variety of options in which only a handful are actually relevant to a user’s specific wants, the user now has to choose from a small list of options that are specifically selected to suit their needs. As such, it becomes significantly harder to actually compare and choose between these options.
Huge’s Sophie Kleber highlights these prediction and recommendation platforms as initial steps in the evolution of anticipatory design. By building algorithms that collect, filter and smartly apply learnings from user-generated data to the product, we’re getting one step closer to products that are capable of taking beneficial decisions for the user.
By focusing our attention on making decisions on behalf of the user, we consciously and deliberately move away from the concept of personalization. We step away from a time where seemingly every content-delivery service seemed focused on serving an "individual", custom-tailored version of the product.
The era of personalization is coming to an end. Enter the era of automation.
As we continue to focus our efforts on building smart, adaptive algorithms, the era of personalization will slowly come to an end. These algorithms will aid us in establishing patterns in individual and collective user behaviour and leveraging the knowledge of those patterns to serve individual users with the right amount of information at the right time.
"Content is king", wrote Bill Gates in 1996. Content, Gates argued, would be where "much of the real money" would be made on the Internet. And right he was. Most digital products stand and fall by the quality of their content. Content is the defining factor by which the quality of a product is assessed. It is why a company like Amazon would rather focus on delivering the most relevant and accurate content than posting a profit.
With anticipatory design however, it can be argued that context, not content, is king. While this doesn't mean that content can be put on the backburner, there is a distinct shift in focus towards contextually aware algorithms.
By continuously analyzing and assessing contextual data, we can assure that whichever decision is taken on the user's behalf and thus, whichever need is anticipated, is actually the right one. Based on this analysis, we can then decide what content we want to deliver.
It is on this basis that digital assistants like Siri, Cortana or Google Assistant and their physical counterparts like Google Home or Amazon Echo are able to deliver the most relevant content to the user. These products are exposed to an extremely large amount of contextual data based on the user’s commands and search queries. In turn, they can learn and adapt to an individual user’s specific needs.
Contextual awareness is also what enabled Google to provide the experience described in the opening paragraph. By interpreting the data available across their own platforms like Inbox and Maps, and third-party platforms like OpenTable, they are able to leverage this immense wealth of information to preempt their user's needs and wants.