Smartphones and the Future
A focus on smartphones excludes the most marginalized
Smartphones are everywhere. They take our pictures, deposit our checks and help us go to bed. To many, smartphones are the future. They’re in the hands of Instagramming middle schoolers. They form the backbone of the burgeoning e-commerce in China. They spark the interest of venture capitalists and revolutionaries alike.
Smartphones — those devices we commonly associate with a touch screen and app store — are not everywhere. Your grandmother may have one, but outside of the context of the Global North, they’re not as common. Even in the fast growing India, only four out of ten own a smartphone, but nine out of ten have some type of mobile phone [1]. Most people use what is called a feature phone — a device from neither Apple nor Android that might look like a brick — or an even more constrained basic phone.
Some rely on the current prevalence [2] to imply that smartphones are the future. This falls short because of three assumptions: (1) it assumes a prior knowledge of which phones classify as smartphones, (2) that phone distribution statistics represent an accurate picture of the world and (3) that past trends in phone distribution will continue. That smartphone ownership has increased in the past does not imply that it will increase in the future. Given that non-smartphones remain prevalent, a focus only on smartphones excludes the most marginalized.
Classification
Phone distribution statistics assume certain classifications: that some phones are smartphones, some are feature phones and none are both. While this classification is helpful, it is also reductive. When funding decisions are made on the basis of such reductions, we suggest a need for clearer classification.
A basic phone can only call and text on the network. Many basic phones are used devices. A feature phone has some apps, possibly runs a version of Java and has a longer battery life than a basic phone. A smart feature phone may not appear different than a feature phone; both have keypad entry and small screens. Smart feature phones — like the much talked about Indian Jio phone [3] — probably have WIFI, a smart-OS, and might have 4G network capabilities. Entry-level smartphones appear different with touch screens, but their battery life or processing power may be no better. They probably run Android and cost slightly more than a smart feature phone. Advanced smartphones have the same attributes as entry-level smartphones, but are faster, have better storage and processing, have a good camera, etc.
Phone classifications based on technical differences go deeper than a divide between smart and feature phones and allow for more nuanced decisions. Even so, as demonstrated in Figure 1, the boundaries between phone classifications remain hazy. Is the presence of a touch screen sufficient to designate a phone as smart? What about WIFI access?
Distribution and Trends
Statistics on phone distribution lack the granularity and verifiability to be useful for all contexts. A much cited 2015 report on global smartphone distribution exemplifies these errors. They asked participants “Is your cellphone a smartphone, such as an iPhone, a Blackberry…” [4]. Such phrasing and country-level granularity reduces data to too narrow a scope. With limited data come limited decisions. With granular data, one could make nuanced decisions based on the distribution of specific device attributes. For example, a payment company might not invest in a camera-based biometric security application if they found few phones with cameras in their deployment area.
In attempting to better validate the distribution of phone types, we could rely on cell tower records instead of surveys as shown in our prior work [5]. Cell towers can store the type allocation code (TAC) of all phones that attempt to connect, thus identifying a phone’s make and model. In turn, this data can be used to lookup phone attributes from a database; we can actually discern if most people in a certain area have a camera phone. Unlike other types of cell records, TACs are more easily accessed from telecoms because the information is not sensitive. TACs suggest a method for empirically determining phone distribution and can become the ground truth for phone classification.
Implications
We imagine three possible futures for phones outside of the Global North: (1) smartphones take over world, (2) smart feature phones (like the Jio) become viable for middle and low-end market segments or (3) the current world continues. Regardless of the technological medium, one will only effectively reach the most marginalized if there is a clear platform for application development. With each future, this would be (1) native Android applications, (2) applications that can run on close to all smart feature phones or (3) the least common denominator between all phones, unstructured supplementary service data (USSD).
Before smartphones can be useful to the most marginalized, they must become more durable (physically and electrically), less expensive (hardware and network cost) and be designed with these groups in mind. The low-footprint Android Go steps in the right direction of smartphones for everyone [6].
More feasibly, feature-like phones will remain prevalent. The Jio Phone, with an app store and easier application development, suggests a middle ground between smartphones and feature phones. It has spread through India largely because of one telecom’s apparent price gouging on 4G data rates [3]. Similarly inexpensive data might allow for the uptake of feature phones in other regions. Even if Jio-like phones dominate the lower portions of the market, interoperability of applications between different operating systems is not certain. If one cannot easily develop an application on all phone platforms simultaneously, the current state of market fragmentation will prevail. In this case, standards for smart feature phone application development might reduce fragmentation. In the worst case, a small number of Jio-like phones is preferable (something like the division between iOS and Android with smartphones).
In the case that low end phones remain prevalent, the most reasonable channel for application distribution is USSD, or unstructured supplementary service data [7]. Using USSD, a user dials a short-code (like *123#) and is served a console-like user interface in which she can navigate through a series of menus by entering commands (as shown in Figure 3). These services suffer from limited functionality available in their text-based user interface. A contract with a telecom, which is required to scale USSD, further complicates and increases the price. If telecoms continue to charge high rates and too closely arbitrate access, USSD will not succeed as a medium. In this future, it is through lobbying telecoms to reduce costs for USSD and to allow more players to enter the space that apps can reach the most marginalized.
While media attention and consumer research offer predictions and exclamations on the distribution of phones, their figures are problematic. Instead, we suggest the adherence to a methodological approach. The truth value of “smartphones are the future” depends on how smartphones are classified, how phone distribution is measured and how one interprets the distribution.
Tim Wu said that “Industry structure…is what determines the freedom of expression in the underlying medium” [8]. The CEO of Safaricom, a telecom in Kenya, has said “We see that the technology we have today is very clumsy. . . We should have open systems. It’s an open world” [9]. These open systems are not guaranteed. For each future, only a series of technological and political advances can engender a platform open to all.
Works Cited
[1] “Consumer Barometer from Google.” https://www.consumerbarometer.com/
[2] James Manyika, Susan Lund, Marc Singer, Olivia White, and James Manyika. Digital Finance for All: Powering Inclusive Growth in Emerging Economies. Technical report.
[3] Order Now Jio Phone — Buy Jio 4G Feature Phone Online for Rs.1500. https://www. jio.com/en-in/book-jio-phone.
[4] Jacob Poushter. Smartphone ownership and internet usage continues to climb in emerging economies. Pew Research Center, 22.
[5] Kushal Shah et al. An Investigation of Phone Upgrades in Remote Community Cellular Networks. In Proceedings of International Conference on Information & Communication Technologies and Development, Lahore, Pakistan, November 2017 (ACM ICTD ’17), 12 pages.
[6] Introducing Android Oreo (Go edition) with the release of Android 8.1. https://www.blog.google/products/android/introducing-android-oreo-go-edition/, 12 2017.
[7] G. Weld et al., “eKichabi: Information Access Through Basic Mobile Phones in Rural Tanzania,” in Proceedings of the 2018 CHI Conference on Human Factors in Computing Systems, New York, NY, USA, 2018, pp. 133:1–133:12.
[8] Tim Wu. The master switch: The rise and fall of information empires. Vintage.
[9] Abdi Latif Dahir. The CEO of Africa’s most mobile innovative company warns his “clumsy” product needs to quickly diversify. https://qz.com/813612/bob-collymore-safaricom-ceo-warns-m-pesa-is-a-clumsy-product-that-needs-to-diversify-or-risk-dying/
