5 Ways Data Science is Changing Logistics Across Africa
By: Jefferson Sankara — Senior Data Scientist at Lori
There are many reasons to get excited about where the logistics space is headed across the African continent. Technology will remake this space, and it will happen quickly. As a data scientist at Lori, I’m looking at new and innovative ways of understanding how to better serve our customers, from cargo owners and clearing and forwarding agents, to transporters and drivers. Lori seamlessly connects cargo to transport, and I’m constantly trying to find data-driven solutions to drive operational excellence for our customers. Here’s a quick overview of what I’m seeing in logistics and adjacent industries:
- Acceleration of Digitization Efforts: The mobility space is going digital. Nairobi has long been hailed as a hub for tech innovation in FinTech, but the tech ecosystem is expanding and new digital tools are changing how complex services are rendered in our region. Widespread adoption of proven techniques like the use of SMS and USSD and incorporation of mobile money transfers into logistics infrastructure is helping solve customer pain-points through the value chain.
- More and Better Data: A key result of the above is more and better structured data available at every stage of the cargo journey (from the loading stage, to on-road journey, to offloading, etc.). For data scientists like me, there is almost a multiplier effect of opportunity to gain new insights into operations once we can connect data sets that have previously been inaccessible.
- Using Data to understand movement of goods: As new analysis can be performed, our ability to create value for clients through more efficient operations becomes a tangible benefit of working with us. Some of this is low hanging fruit, like connecting publicly available data from ports to our own proprietary data from customers. Other times it’s a matter of getting creative with third parties with data on, say, geolocation and journey time related information, like the Uber movement platform provides.
We’re seeing more and more companies use data science in the areas below; these areas are important to Lori as we deliver on our mission to optimize operations and bring down the cost of goods in Africa. Here are some key trends:
- Data Science for Better Customer Understanding
Organizations in the logistics space are using data science to understand their customers better. This is currently being done through the analysis of customer data to determine and measure metrics like conversion and churn rates, customer loyalty, customer product/service profiles etc. In our space, we can quickly generate customer profiles based on trip data on loading times, offloading times, and even fuel consumption. There are success stories from the FinTech and Telecom industries that logistics service providers can learn from too. Safaricom is a good place to start. They are using data to churn out new products which are cross-sold to their customer base. A great recent example is the Fuliza M-Pesa overdraft service. You can read more about it here.
2. Data Science for Quality Stakeholder Exploration & Discovery
Organizations in this industry use data and data science tools and methods to analyze data and understand the stakeholders of their business. These include; drivers, suppliers, governmental and local authorities. Using data scraping techniques, they are collecting data from their platforms and other platforms like social media and then applying techniques like sentiment analysis to find out what the stakeholders think about them and what strategic or operational steps they can take to improve their services and their products. A great case study to learn from is the Repustate; a Multilingual Text Analytics for Businesses platform that helped a South African Bank to understand their customers perceptions of their services. The case study is well detailed here.
3. Data Science for Market Analysis and Customer Segmentation
Customer segmentation or audience segmentation (the practice of identifying common characteristics of portions of your total customer base) helps companies build better products that meet the exact needs of groups of customers. Data science helps build out profiles of customer types, which product teams like ours use to help develop new tools, products, services, and features for our most important customers. Common tools we use include statistical analysis and machine learning on large data sets to generate business intelligence and insights. Doing so helps us make better decisions from product, packaging, pricing and beyond, while serving our customers changing business needs.
4. Data Science for Asset Tracking
Logistics asset tracking is already a well established business model in the African continent. From the tracking of vehicles, trucks with cargo and cargo on maritime vessels, companies have invested in both software and hardware to understand what’s going on with a particular asset at any stage of the delivery and return journey. The best way to perform analysis today is by using data science and data visualization techniques. This allows one to better understand the relationship and context of a piece of information. For instance: if a taxi driver claims that a trip from Nairobi’s CBD to JKIA airport on average takes about 45 minutes, this information has less context to it and it would be misleading without the understanding of which day of the week it is, what time of day it is, what the weather condition is for instance (and this will seem obvious to anyone who experiences Nairobi traffic.) The implication is that proper trip planning and route decision making can only take place with the right contextual information. The better that context is, the easier it becomes to guarantee your passenger will make their flight on time.
5. Data Science to Measure (and then manage!) Key Operational Metrics
Organizations are using data science techniques to measure key operational metrics regarding their trips, specifically as it relates to journey milestones that take time and have cost implications (loading and offloading times, border crossings, etc.) Managers and decision makers want to understand trip turnarounds, and be able to drill down to specific stakeholders like drivers and vehicles. They perform analysis to understand their performance in various routes. The data is only meaningful when it is specific enough to generate insights and/draw conclusions about the source of delays, etc. At Lori we derive these insights from operational data, analyse the same and help our clients understand their business better through our flagship: the Lori Client Application (LCA).
Overall Outlook and Key Takeaways
As data science is applied to the logistics space, there will be increased efficiencies in logistics operations across the continent — that is almost a given. For Lori, this practice directly supports the achievement of our mission: to reduce the cost of goods in frontier markets.
One pervasive challenge remains however: transforming this industry’s talent supply. For example: I am a more traditional programmer by training with a background in mathematics and statistical analysis and I have made the switch to Data Science. I am encouraged by efforts in the market to train and groom talent, and I am always looking to engage, mentor, and help groom the next generation of home-grown Kenyan talent at Lori. Lori’s in-house, Kenyan tech team cultivates local leadership in this space. If you’re excited about where the industry is headed and the opportunities Lori is opening up across Africa, join me on this amazing journey!
Jefferson Sankara is a Senior Data Scientist at Lori Systems and focuses on machine learning. Prior to joining Lori, Jefferson was a software developer at Trade and Development Bank. Areas of expertise include Data Management, Business Analytics, Software Development and Business Continuity Planning. Jefferson earned an MBA in Strategic Planning from Edinburgh Business School (Heriot University) and holds a BS in Applied Computer Technology from USIU — Africa.