Why don’t mobile money providers own their product?
Friday, September 14, 2012 at 2:06PM
Neil Davidson in Mobile Money

[Quick note: We have a shiny new comment management system (Disqus) and hope you will use it. Unfortunately, with apologies to Ali, Michael, and Harold, the comments made on our old system were lost in the switchover.]

Paul and I are the founders of a mobile payments company in Southeast Asia. Long before we started, we decided that we wanted to build our own transaction-processing platform rather than license one from a vendor. It’s a decision that has had profound implications for our company. And it got me wondering why it is the case that the majority of mobile money deployments use off-the-shelf transactional platforms.

Traditionally, when it comes to software, you buy it if your requirements are standard and stable and you build it if they aren’t. This is why Silicon Valley startups build rather than buy: Mark Zuckerberg never had the option to buy an off-the-shelf social networking platform. Conversely, this is why banks buy rather than build: banks tend to want the same things from core banking systems, so it’s more efficient to buy them from specialized vendors rather than build for scratch.

Vendors in the mobile money world did a good job of persuading mobile operators that their needs were predictable and standard, implying that it would make more sense to buy rather than build. They were helped by the fact that mobile operators tend not to have strong software engineering or product management capabilities in-house. Off the top of my head, I can think of only one MNO, and just a handful of third-party services, that have build their own platforms from scratch.

Increasingly I wonder if this was not a mistake. Building your own platform allows you to take ownership of it in a way that is otherwise impossible. In Silicon Valley, startups are taught to constantly iterate their products in order to attract more users and engage them more deeply. This is impossible to do if you have licensed a platform from a third party. At MMU I talked to countless mobile money managers who were banging their heads against the wall because even the simplest “change requests” would take months and tens of thousands of dollars for a vendor to implement—not exactly conducive to rapid prototyping.

It’s become increasingly clear that for mobile money to succeed, it must become more responsive to the wants and needs of customers, which increasingly seem to vary significantly across markets. The idea that M-PESA clones would succeed everywhere has now been thoroughly debunked; yet most mobile money providers are still running transaction-processing platforms that essentially replicate the functionality of M-PESA. This suggests to me that the adoption of standard mobile money platforms around the world was, at best, premature.

Article originally appeared on Insufficient Balance (http://insufficientbalance.com/).
See website for complete article licensing information.