A network effects business model allows a company to gain more value as more companies use its products or services. The value of the offering rapidly increases because each additional user increases the value of the network. Today, we mainly think of digital companies like Google and Twitter in this category. Google and Twitter mainly monetize the data through targeted advertising. The value of the data captured by Google and Twitter has made them the darlings of Wall Street.
But it turns out big logistics firms also generate Big Data, and they are also working to monetize this data. A case in point is FedEx. FedEx Dataworks is an organization within FedEx that uses data science and machine learning to help make shipping more efficient. The richness of FedEx’s data comes from the 17 million packages they ship per day around the globe. Each shipment is scanned many times as it moves from origin to destination.
Sriram Krishnasamy, the CEO of FedEx Dataworks, said that the data is much richer than just scan data. FedEx enriches this data with weather and traffic data. FedEx is doing the obvious thing with that data; they are using predictive analytics to improve the flow of goods through their network and make themselves into a more reliable carrier. “Information is just as important as the package,” Mr. Krishnasamy asserted. “We have always been good at capturing data. Now it is in a big data infrastructure” and they can find connections between data that may look unconnected.
This predictive intelligence is particularly important for high value packages that must be delivered quickly. A case in point is the FedEx Surround solution. FedEx Surround is based on proactive monitoring and intervention controls across a delivery network. FedEx Surround predicts shipment success by combining information about a package with external data such as weather to manage the risks surrounding the shipping process. Shipment data includes scans, shipment route, conveyance and SenseAware ID, a small Bluetooth tracking sensor that can be placed on a package. This allows a white glove service team to monitor the location of a package and take action if it looks like the package will not arrive on time. For example, an intervention might consist of a team scooping up the packages at risk at a sort center and putting them in a van that will leave immediately. But, from origin to destination, FedEx asserts, there are multiple intervention points.
FedEx Surround’s first use was in tracking critical, sensitive shipments of COVID-19 vaccines as they moved from manufacturers and distributors to medical centers across the U.S. In the first year of COVID-19 vaccine distribution, FedEx delivered approximately 300 million doses throughout the U.S. with the help of predictive analytic tools. The vaccine shipments move through the FedEx network with an average transit time of less than 20 hours. The tracking devices, combined with the white globe teams monitoring the shipments, allowed for proactive interventions that led to a 99.91 success rate on getting vaccines delivered on the day of commit.
But FedEx Dataworks is looking to do more than just improve their ability to be a reliable partner. And they are looking to leverage their customers’ data, not just their own. Currently, they have proof of concept project using the data of one of their largest customers. Customer data is being ingested through an API framework and then combined with the FedEx network data.
In one use case, customer data is being used to help predict demand and returns in a joint planning process. The better the forecast around demand and returns, the more the customer stands to save on shipping by injecting the package at the right time and place into the FedEx delivery network. For example, seven-day ground is cheaper than express deliveries. The right injection point is not necessarily the FedEx sort center closest to the customer. Traffic and congestion might suggest injecting packages into a more distant sort center will actually improve the odds of an on-time delivery.
Over time, machine learning will be leveraged so that forecasts will get better. FedEx will also learn from its failures. To the extent that the joint forecast helps inventory to be positioned in the right location and shipped using lower cost services, both FedEx and their customers can benefit. The customer benefits by paying for the service that best meets their needs. FedEx benefits from stickier customer relationships.
There are other opportunities to monetize the data. FedEx ships to 220 nations. Trade documentation is needed for international shipments. Properly classifying goods for duties and insuring goods are not shipped to bad actors (denied party screening) may sound mundane. It is devilishly tricky. Machine learning based on Big Data can help FedEx serve their customers by getting better and better at this.
Another opportunity is fraud prevention. FedEx has validated addresses in their databases. This might create an opportunity to expose fraudulent transactions.
In conclusion, network effect business advantages are not something that accrue only to large digital companies. Big carriers have opportunities not just to improve their service, but to monetize this data in a variety of ways.