Make a Well-Conceived Plan to Ensure Delivery Efficiency: Identify North Star Metrics and Target High-Quality Vendors
Package delivery is the key to the proper functioning of e-commerce. However, in real-life scenarios, express delivery may ultimately become “slow delivery”.
For retailers, this also indicates an increase in negative reviews, a decline in sales, a backlog of inventory, and an increase in operating costs. For these retailers, if they can increase the delivery efficiency based on their data analytic insights, it would mean a lot. In this article, we use the delivery truck trip data of Ram Thiagu to demonstrate how the delivery efficiency can be increased by 13.65% with 30 minutes of data analysis in Kyligence Zen.
The Metrics for Logistics On-Time Delivery Analysis have been published in Kyligence Zen Metrics Template market. You may click the link and reproduce the use case in your own Kyligence Zen account.
First, we will upload the upload CSV in Kyligence Zen. In the Data tab, click + New and then drag the file to the upload area. Upon the completion of the import, you can check the total columns and rows of the csv file and preview the first ten rows of the table. After that, you can go to the Metrics page, and click on + New > Import Metrics to create the logistic-related metrics in one click. Below are the metrics we created:
The Metrics page of Kyligence Zen is now as shown below. It can be seen that the enterprise has delivered a total of 6,740 orders, among which 4,250 were delivered on time, with an on-time rate of 62.94%.
Click the metrics card to enter the metrics details page, and then we can study the impact of various factors on logistics and distribution by changing dimensions. For instance, we can check on-time rates from different perspectives (e.g., GPS vendor, long-term vendor and product category). Meanwhile, we can also switch between column charts, line charts, or pie charts to visualize and compute analytical data in an optimized approach.
Additionally, on the Definition tab, you can view the metrics definition to see which data source the metric is based on and how it is calculated. In case of inter-departmental collaborations (e.g., when you need to communicate with a vendor about logistics and distribution), it is easy for all parties involved to reach a consensus on relevant data, without having to repeatedly confirm the data benchmarks.
Let’s return to the issue at the beginning of this article. Which delivery partners are actually slowing down our express delivery? To this end, we’ll draw a chart to analyze the number of delivery by partners. From the chart, we can see that there are currently a total of 24 vendors, 6 of which made less than 50 deliveries, which made a limited contribution to the business. Therefore, let’s remove them from the vendor list. Then we’ll superimpose the number of on-time deliveries on top of it and sort again. It can be seen that some partners, such as SRTEXKOR96, have very high delay rates. SRTEXKOR96 delivered a total of 172 orders, among which 171 were delayed. Let’s remove these partners from the vendor list. Then we have 10 vendors on the list as premium vendors. The chart below shows the ten vendors we screened out, their total number of deliveries, and their on-time rate.
As the first shot fired in the Delivery Efficiency Defense Battle, we used Kyligence Zen to screen the enterprise’s current delivery partners. From the previous list containing 24 vendors, we removed vendors with low delivery volume and high delay rates, thus shortening the vendor list to 10 vendors, which have undertaken 75.52% (5,090 pieces) of the enterprise’s current express deliveries tasks. As a result, the average on-time rate has increased to 76.59%, up 13.65% compared to the previous period. While improving the enterprise’s logistics on-time rate, maintaining vendor relationships has also become more straightforward. As the entire analytics process is fair and transparent, and the analytics results are more convincing. There is no longer a need to repeatedly align metrics definitions when conducting cross-team or cross-departmental discussions.
Kyligence Zen also has a low technical threshold. The traditional model development approach requires developing a separate data pipeline for each use case, before building the data model. With Kyligence Zen, we can try from all angles with only four metrics, and get the metrics data we actually need. With a low technical threshold, Kyligence Zen can mobilize business users to explore metrics for their own business purpose. Meanwhile, thanks to the one-stop metrics store provided by Kyligence Zen, scenario-specific exploration by business users can also be saved as standard metrics for the enterprise, thus accelerating the accumulation of data assets.
Below are the datasets and North Star Metrics definition files used in this analysis. Please feel free to visit Kyligence Zen website to try them out. Should you have any questions or find a factor that more seriously affects the logistics efficiency of this e-commerce, please leave a comment!
The Metrics for Logistics On-Time Delivery Analysis have been published in Kyligence Zen Metrics Template market. You may click the link and reproduce the use case in your own Kyligence Zen account.
