The Age of Aggregation
Surviving Information Overload
It has been said ever since the time of the Library of Alexandria, mankind is producing more information than could ever be recorded and properly documented. However, for the longest time, we were facing problems of information scarcity. Old forms of media such as books and scrolls required time-intensive manual labor to reproduce. The high cost of reproduction in combination with very low literacy rates among the people limited the spread of knowledge through printed material. Furthermore, before the invention of telecommunication technology, we relied on couriers and pigeons to transport information, often delaying the decision-making process. But the proliferation of different media formats, all the way from book printing, to electronic media, and eventually to the internet, has increasingly helped to solve the problem of information scarcity.
Facing the Flood
Today, we have an overabundance of consumable content and businesses have to be evermore intentional to make sure the content they are choosing to share with their users is valuable. With a large variety of content creation tools becoming openly available and easier than ever to use, more and more users turn into prosumers regularly shifting between consumption and production. Being exposed to an environment so rich in information causes a serious strain on our attention. We have moved from information scarcity to information overload, and thus attention scarcity. Attention is a zero-sum game; we only have so much of it available, and hence need to be careful what we use it for.
With limitless amounts of content available at our fingertips, there is a rising need for organization. Curation plays a necessary function of sub-inventories for what is available, since the sheer amount of what is actually available is unsurmountable.
In the model of traditional media, we had a system of gatekeepers in place who selected from a variety of sources and curated the content that was presented to us through a limited selection of channels, via radio, newspapers, or television. While this system had its own challenges and problems, such as its susceptibility to agenda setting, it presented a mechanism that helped us cope with an ever increasing flood of information.
With the dawn of the modern internet, these mechanisms no longer suffice. The sheer amount of information available far surpasses the capabilities of traditional curators to cover a broad range. In addition, the constant stream of new content also requires a quick processing, in order to not miss out on important new developments — a problem that becomes especially apparent in news reporting.
More and more, the job of filtering through the flood of potentially meaningful bits and pieces is outsourced to aggregators. Aggregators algorithmically sort through a multitude and variety of streams and channels and present the user with a boiled down overview. We see an increase in dashboard applications for a variety of purposes, from giving us a summary of our financial activities like Mint, to applications like Apple’s Health on iOS which aggregates health-related information of its user.
With the constant increment in available options that demand our attention, this service of filtering and doing the leg work for users is becoming a significant value proposition. If attention is becoming an increasingly valuable resource, anything that helps us save this resource also becomes more valuable.
In recent years, we saw the development of many services that tap into this need for filtering through vast amounts of options within defined spaces.
Pandora has made a name for itself as a service that helps its listeners discover new music, based on constantly fine-tuned filters. It dramatically facilitates the music discovery process by letting users set parameters that define the type of music they are interested in hearing at a given time. But instead of presenting users with all the available options, and possibly causing choice paralysis, Pandora plays users one song at a time. It then allows the users to provide feedback with very simple options in order to adjust the selection filters for upcoming music.
Spotify has recently improved its own music service by including a variety of aggregated playlists that helps its users navigate its massive library of available music, especially its Discover Weekly playlist. Based on the user’s previous music history, it proves to be a powerful tool to unearth new tunes.
Similar developments can be observed in the news space. For years, the news app Flipboard ranked among the more popular applications on the App Store. It lets users select from a range of topics that interest them and presents articles in a layout that is easy and engaging to navigate. More recently, Apple introduced Apple News on iOS, which takes a similar approach, letting users select topics and sources of interest in an aggregated and digestible format.
In a wider sense, even services like AirBnb and Uber can be regarded as aggregators of offerings in their respective niches. Neither of these services owns the actual product or content, but bases its business on the service experience around the handling of large amounts of data and helping users maximize the value they can draw from this information.
The Rise of User Curated Content
Furthermore, we see a proliferation of services that allow users to be curators in their own right. The poster children of this development are tools like Pinterest, the visual bookmarking tool, and Tumblr, the microblogging and social networking tool. Both of these let users curate content on their own accounts, as well as through following other users whose content appears in long streams.
A very similar case can be made for platforms like Twitter. Its platform enables the growth of asynchronous networks in which users curate a list of sources — individuals, brands, or organizations — who then serve as curators of content. Whereas tools like Tumblr or Pinterest include the actual content, Twitter works as a tool for directional curation; it is a way-finding system for curated content. Maria Popova, editor of Brain Pickings, therefore describes “Twitter as a conduit of discovery and direction for what is meaningful, interesting and relevant in the world”. Its strength lies in the brevity of the individual pieces of content, that are quick and easy to scan, and only serve as directional pointers, instead of including the actual information itself.
The inclusion of curatorial functionality can be seen in a majority of modern applications, such as collections on Product Hunt, playlists on music services, or even groups in social networks. A case can be made that curation itself is a core principle of modern web applications based on user-generated content. Much of the content shared online is a reconfiguration of previously existing material, modified in some aspect and presented in different contexts.
New Landscape, New Challenges
This development in the digital services landscape brings a couple of interesting challenges with it. Applications need to ensure that relevant context is not removed in process of abbreviation. Our increased reliance on filters requires a careful adjusting of the parameters in order to avoid the creation of so-called filter bubbles. And how can services build a relationship with customers if their content is largely consumed via middlemen, either curators or aggregators.
Despite major advances in natural language processing, many services rediscover the value of human taste in curatorial efforts. A very recent example are playlists compiled by guest curators in the newly launched Apple Music service. It also shows the role influencers can still play as tastemakers and gatekeepers even in today’s hyper-automated services. However, relying on manual curation is a risky business.
Most curatorial work consists of sourcing content from external sources, with the act of curating being an important creative labor. In this construct, the curator or aggregator plays the role of a middleman between consumer and content creator. Because of this constellation, the ownership of the experience becomes increasingly complex. While users perceive the value of the service, they use these tools for the presented content, which is often times outside of the control of the platform provider. The relationship between content creator and platform provider is thus highly interdependent. If the user experience of consuming content on a specific platform is significantly superior, users might perceive the presented content sources as exchangeable, which presents a severe risk for brand loyalty.
Nowadays, our modern web platforms serve as enablers of conversations. But these conversation are largely outside of the influence of content creators such as brands or institutions. Aggregation and truncation of content leads to a massive decentralization, which makes it very difficult to participate in or have access to the core of the conversation at hand.
In addition, the proliferation of content creation tools renders it nearly impossible to monetize content alone. Unless a platform can provide exclusive unique content, it is extremely difficult to monetize this content, because the sheer amount of free and openly available content presents the user with a myriad of alternatives.
Therefore, providing users with an excellent experience is the most important value proposition platform providers have. One example for this is Medium, a publishing platform that prides itself in providing a superior reading and content consumption experience. While some of its content can be found elsewhere, its service is building a loyal following through the high quality of its experience.
But even with carefully crafted experiences, combined with curatorial efforts, this space is still searching for suitable business models. Circa, a news app that featured manually curated news content presented in small, easily digestible chunks, closed its service because it was unable to build a sustainable revenue stream. Despite its well-executed service, it proved that users were not willing to pay a substantial enough amount to maintain the business. Additionally, these manual efforts are enormously difficult to scale, which presents a further roadblock for long term monetization and growth.
Another interesting challenge that results from the growth of aggregation as a core service is the flexibility required with regards to interface design. Tools need to be able to handle a large variety of different content forms and media. Over the past few years, cards have become a popular solution, as seen in tools like Google Now, or even Pinterest. They can be flexibly arranged to accommodate for the wide range of different devices on which content is consumed, from large TV screens, to tablets, smartphones, and even wearables. In addition, they present self-contained units, which serves as a powerful symbol for the truncated nature of aggregated information.
And while behemoths like Facebook, in their complexity comparable to operating systems, try to provide solutions to navigate an ever increasing array of media formats, we see developments on the opposite end of the spectrum as well. Hyper-focused applications are introduced, focusing on narrowly defined formats or tasks. Instagram was introduced as a single-purpose application for sharing photos. The Facebook messenger was spun off as its own application, before eventually evolving into an ecosystem in its own right.
The example of Facebook Messenger also serves as an example for another promising approach to the challenges in the age of curation. Messenger and several other applications provide a minimal, text-based interface for a variety of tasks. They allow language-based interactions with data, much like conversations with a human service provider. Facebook takes this approach further with its project M, a text-driven virtual assistant. Other applications try providing this type of service for much narrower scopes, such as Penny or Operator.
These are the fundamental challenges we face in this age of aggregation.
Content creators need to ask how they intend to differentiate themselves in order to create customer loyalty. Some may attempt to build and maintain their own services, while others might try to create a walled garden around their content — such as paywalls on news sites — to encourage customer loyalty through a sense of exclusiveness and artificial scarcity.
Platform providers have to find ways to balance ease of access and truncation, while keeping relevant context available, an aspect that is especially challenging for purely algorithmic aggregation. The processing of natural language plays an important role in this area.
Designers need to develop solutions that can handle the increasing complexity by including features like templated solutions with customization options. Additionally, these new services might require new forms of interaction patterns, especially with the emergence of new types of consumption devices.
The modern internet is still in its infancy. There are no definitive answers on how to solve the aforementioned challenges. Whether dashboards or No UI solutions, human curation or algorithmic aggregation, or combinations of different approaches are most appropriate will greatly differ based on the context in which services are used.
While advances in big data and natural language processing will play a big role in shaping the development of new service offerings, the experience needs to be informed by an empathic understanding of user needs. Keeping the user’s goals in mind when crafting new services will help identify when complexity needs to be hidden, and in what contexts to provide more complex options. Moreover, identifying the user’s emotional needs can guide decisions about integrating human service providers into otherwise automated processes.
Thus, while major technological advances will allow companies to offer new kinds of products and services, these offerings need to be designed around the users’ needs and motivations, helping them achieve their goals.
This article was originally published on the Universal Mind blog.