Wednesday, December 7, 2011
Liz Losh's Guest Blog: Design is More Than Copy and Paste
In the final panel of the annual IMFTI conference on "Mobile Money: Lessons for Microfinance and Design" moderator Paul Dourish emphasized how the shift in thinking from "user-centered design" to "design-centered use" could be applied to mobile money systems. (The video above shows mobile phone operated vending machines in China.)
The first talk on the "Impact of Mobile Money Services on Microfinance Institutions by Patricia Pulido, Maricruz LaCalle, and Casey Conzett focused on an analysis of operational costs in Tanzania, where they said mobile money services were not developed despite the entry of companies like TIGO and Vodacom in the market and a growing role for microfinance institutions that emphasize small-scale finance. Researchers studied 37 institutions all over the region, which included places like Mufundi Community Bank, Njombe Community Bank, or Tandahimba Community Bank. They interviewed general managers, loan officers, and other bank personnel and identified a number of reasons that mobile money might be appealing, including "flexibility to adapt to client needs" (44%), "greater outreach to rural areas" (30%), "new sources of revenue by commissions" (30%), and a "greater number of clients" (15%). They also cited difficulties when electricity was unstable or the network slow, problems with training and marketing existed, or forced loyalties restrained consumer choice. Although they lamented limitations of time and acknowledged that "the field is not a laboratory," they pointed out that this was also highly original research given current literature reviews.
Speakers Panthea Lee and Zack Brisson of Reboot presented a design-savvy talk about "Value Systems in China: A User-Centered Approach to Designing Inclusive Second-Generation Banking." (Brisson was formerly active in the anti-genocide organization Enough.) In leading off his talk, Brisson joked that it was "almost as scary talking to a room full of anthropologists as it is to talk to a room of psychoanalysts." He described ReBoot as devoted to "getting to better research results" by understanding "lives and contexts" and the fact that "process maters." In focusing on China as a region of interest with an economy that was "literally skyrocketing," he argued that it was important not to forget the "many left behind" by "exacerbated economic inequality. Now, he insisted "innovation is possible" if it aims at "an inclusive, second-generation banking system. Brisson argued that China was a particularly likely area in which the unbanked could become banked through mobile money, because 70% of the population used mobile phones, remittances already shaped financial practices, and there were many existing agents. Like M-PESA in Kenya, there was also a clear vision for partnership between the telcom and the financial institution, as in the case of China Mobile and Shanghai Pudong Development Bank. Although these technologies may follow different uptake patterns, "desire for financial stability is universal."
Brisson and Lee described the core of their methodology as ethnography. Their study of study of mobile money focused on 4 cities and 6 towns and villages across a cross-section of the country. During a period of three weeks they conducted 113 interviews in the context of homes, places of business, or communal gathering points. The study relied on both unstructured and structured interviews and observations of direct service use in sectors such as health care and travel. As they explained, after unprofitable rural banks were closed as a result of market reforms, there was a strong need for mobile money to improve access to financial services. They noted that migrant populations were another important factor in the economies that they studied and that such Chinese often treated as another commodity, although they contributed to half of the nation's GDP and worked much needed high-risk, low security jobs. Furthermore, the marginalization of minority populations could create more obstacles to financial inclusion for certain segments of the unbanked, such as herdsman who had been forbidden from raising livestock by the authorities. (Older parents may have been compensated for the loss of their livelihoods, but children were not eligible for these government payments.)
Lee and Brisson emphasized three main themes:
1. Trust, defined by the in-group tendencies of kin, caste, or geography
2. Uncertainty, which might cause some to choose to forgo the risks associated with pursuing wealth. (As one couple said of their lives under Mao, "We knew we were poor, and would continue to be poor. That's better than not knowing where we will be tomorrow.")
Lee and Brisson argued that it was necessary to design for agents, intermediaries, and influencers, not just end-users and to focus on trust, networks, and relationships. (Those interested in this approach may also want to check out the "infomediaries" research of The Global Impact Study and the work of François Bar.) For more about the work of Lee and Brisson in China, see ReBoot's blog posting on "Mobile Money in the Land of Mao."
"Best Practices in Mobile Microfinance" by Fatima Yousif, Elizabeth Berthe, and Olga Morawczynski provided a global overview of how the technology was being adopted in multiple countries. As Yousif explained, although MNOs focus on quickly profitable services, MFIs have to focus on the hard-to-reach and difficult-to-profit from. In their study an online survey was sent to over 100 microfinance institutions, and the group also conducted direct interviews of 16 MFIs, mostly in Kenya 16 MFIs, with the aim of addressing the relationship between "industry and us" and the "need to address real needs." By focusing on areas like low agent penetration, the group was able to examine social and commercial sustainability. Socio-cultural complications and research challenges are inevitable when mobile phones are frequently shared, there are problems with checks and controls, and there are low literacy levels. Yousif noted that in Cambodia low-end phones might not able to read Khmer script, and that there may be other complications in countries with citizens who have low literacy levels in own language to begin with relying on text applications. She also observed that there may be situations in which it is still less expensive to travel to a bank than to use mobile payment services, so newer technical solutions weren't always welcome. She said that it was surprising to see how few institutions did cost-benefit analyses, and how many providers assumed that there was no need for market research, because a particular approach was "obvious." In a market dominated by money transfers, where mobile network operaters know they can make the most profit, innovation was often hampered by pre-existing assumptions and the fears of stakeholders. For example, loan officers feared losing their jobs if new technologies were adopted. She also pointed out that the success of M-PESA was "both good thing and a bad thing," and she cautioned against "copy and paste" approaches in other regions. Certain factors may be "necessary but not sufficient," particularly in markets that are urban and peri-urban. She argued that "IT/MIS integration is one of the greatest challenges faced by MFIs today," a problem exacerbated by top-down decision-making.
She closed with a number of recommendations, which included "understand your market environment," "communicate, communicate, communicate," "invest time in developing your distribution network," and "test and monitor your product," because "new uses will come up." Her recommendation to "collaborate with regulators" proved to be the most controversial piece of advice during the question and answer session that followed.
As moderator Dourish asserted, "HCI not about interfaces but about relationships between design and use" and "producing designed effects." In answering his own question about where the sites of innovation might be, he emphasized the importance of "temporalities of innovation" and "directionalities of innovation." He also noted that "operator selection" was a "design decision in its own right."