Mass Transit seems to be an unsolvable problem in emerging markets.

Every city has its inglorious stories of informal transport struggles and endless traffic jams. Every urban citizen from Nairobi to Dakar will also recognize their own city. Public transport operators have failed and a plethora of informal buses, minibuses, shared cabs and motorbikes have taken over without fully meeting an ever growing need for transport.

Depending on whom you ask, the solution could be one of a wide variety of silver bullets: more regulation, more roads and infrastructure, bus rapid transit, light rail, GPS tracking, and cashless payment. A couple of years back, some players (including Safaricom, Equity Bank and Google) engaged in a battle against cash in the mass transit sector in Kenya. Despite the incredible penetration of mobile money and strong government support, the 'cashless matatu' initiative did not really take off.

The trouble with typical suggestions, is they ignore one of the most critical player in this ecosystem - Matatu drivers.

The Matatu Driver is the most critical player

The matatu driver is very often the villain in every bad ride experience story or road incident news. They are portrayed as being irrational to the point of incorrigibility, but strangely, nobody has tried to understand the root cause of their behaviour. Why do they keep riders waiting for hours for the matatu to get full before departing? Why do they stop every few miles to pick passengers along the road or force riders into another bus, mid-journey? Why do they race between some stops, at the risk of causing accidents?

The reality is that drivers are "blind" when it comes to assessing demand. From a given stop, they have no idea about the number of riders in all subsequent stops along the corridor. For drivers the trade off is to sacrifice passenger experience in an attempt to maximize their returns per trip, and consequently, their daily gain.

Before adding more roads, regulations or cashless payments, mass transit operators in Africa should start by using technology to track demand (time, location, travel class) so that drivers can more efficiently target riders. This leads to less random hops, less wait time for riders, less fuel consumption and better filling rates.

At any given time, the primary need for a Matatu driver is to spot the top stops with the largest number of passengers going to a selected destination, and with this they can maximize filling rate and trip revenues.

Looking at this analogy, one can see certain similarities with Search Engines. For a given topic (keyword), they are able to find the top sites with the most relevant and timely content online.

Lessons for Matatus from Google's Search Engine

Today's search engines look almost the same - a blank box on any browser that provides speedy and relevant answers to queries, with a good success rate. A lot has been written about Google, but no link has yet been established between web search and mass transit.

The "portal wars" of the 90s between Yahoo, Lycos, AOL and AskJeeves did not consider search as a key weapon. All what mattered was keeping web users in one portal, and feeding them with private content (almost like social networks today). Displaying search results and links that routed them out was out of question.

As the sprawling web continued its exponential growth, finding accurate results started to become a major issue. Many portals such as Yahoo! started building curated directories, organized in themes to cover more and more sectors. When you searched on Yahoo!, hundreds or thousands of answers would be displayed in no particular order. Additionally, the directory was very far from being comprehensive, some search engines could not even find their own names as a keyword).

What was later to become Google, started from an attempt to search the entire web, not just private directories, and display the results with some priorities – relevance dimension.

In scientific research and publications, the importance of a paper is directly derived from its citations - the number of times the paper has been cited by other scientist papers. Google mimicked this approach and begun ranking each website based on the number of other websites linking back to it. This is how PageRank started, along with BackRub, the search engine accidentally derived from this method. The rest is history!

It is actually straightforward to apply the BackRub approach offline, to mass-transit in the real world. Just like the web, Nairobi, Lagos and Dakar are a dynamic collection of millions of places (hotspots) where riders flock to, at any given time. The "importance" of a place is directly related to the number of trips (thus passengers) heading to it.

This is why a special area in the city, like a shopping mall, suddenly becomes trendy. Just like web back-links, for a given place: a school, a stadium, or an airport, it is possible to determine the top departing places that are "feeding" it in traffic. With this handy data, matatu drivers could easily target those pick up places, in priority.

For a given destination, the matatu drivers can simply get the best hotspots to directly pick passengers up, instead of guessing and annoying their clients. After all, this is the primary reason why taxi drivers are paying 20% to Uber, in the on-demand ride hailing sector, for the ability to find riders quickly, at proximity vs randomly wandering across the streets to find riders.

In reality, transport has its own complexities: as an example, the hotspot's ranking varies over the day if not impacted by hyperlocal events such as incidents, presence of police, or whether the traffic lights are working in that particular area. In general, however, the chaotic mass transit in emerging markets can be solved with a dispatch inspired from Search Engines and the original ranking system, itself derived from scientific paper citations.

Africa's urban planning regulations, road infrastructure design, real-estate development and mass-transit dispatch should actually be redesigned from the ground up, incorporating this demand data to keep track of where passengers are, and enabling drivers to go where the need is rather than leave it all to chance.

This post originally appeared on Matatu Mobility

Share this via: