Big Data & Logistics (Part II): How to implement?

In our previous post titled “Big Data & Logistics (Part I): Breaking Down Big Data“, we covered the what and the why of Big Data.

The what – We demystified the term Big Data as the analysis of multiple giant data sets to reveal relevant patterns and trends. The main problem that companies face when dealing with Big Data is the incompatibility of the companies’ existing infrastructure to crunch the sheer volume of data.

The why – We determined that the value of Big Data lies in the integration of data analytics with companies through operational efficiency and customer satisfaction. Companies stand to increase their level of operational efficiency through an increase in last mile efficiency, making relevant changes following the tracking of various KPIs, as well as leading companies against industrial benchmarks to maintain relevance and competitiveness. Also, companies can improve customer satisfaction through identifying key drivers of customer retention, understanding of trends in transactional data and analysis of customer feedback.

In sum, logistics companies stand to lose or lag behind if they do not integrate data analytics. The Boston Consulting Group highlights this through its December 2016 report, which predicted that l

ast-mile costs stand to fall by more than half if business strategies based on data analytics pay off.

In this post, we look at the how of Big Data. Specifically, we explore the ways and processes of implementing structure in companies in order to reap the benefits of Big Data analysis.

What to measure?

Gather your stakeholders

Before starting the actual integration of data analytics for the company, it is imperative to be aware of the operating environment. Specifically, gather and identify the different stakeholders contributing to the business in order to build a clearer picture of the current logistics network. This process involves identification of the drivers, 3PL players, 4PL players, customers and the consignee.

For this article, we will be focusing mainly on the transporters and what they can do to implement Big Data. This is not exclusive for transporters though and can be applied to other parties in the supply chain.

  • For the brand owners, they are the main source of delivery orders. They own the goods and requires their goods to be delivered. Understanding the nature of their business, be it e-commerce, manufacturing, retail distribution or import/export and the kind of delivery type required, allows the company to configure their operations to the brand owner’s needs.
  • For the forwarders, they are the intermediaries between the orders gathered from the brand owners and the transporter. The forwarders are the one who has won the logistics contract with the brand owner and are responsible for coordinating and finding the transporters with the right expertise for the jobs. Understanding the style in which these forwarders operate and where the bulk of the value of orders lie in, helps to capitalize on the potential profits.
  • For the drivers, they are the backbone of the industry and have the power to determine the success of delivery operations as they are the ones on the ground fulfilling the delivery orders. Understanding the size of the company’s delivery fleet and what makes them tick allows a better allocation of resources.
  • Lastly, for your consignees, they are the end-recipients of every delivery. Understanding their expectations is key to ensuring that every delivery is a successful delivery.

Generating KPIs to measure performance

After understanding the players in the operating environment, it is crucial to define several Key Performance Indicators (KPIs) for performance tracking. What the company is measuring will play a big role in determining what tool to choose and how to go about collecting the data required.

The following KPIs are industry benchmarks which are recommended as the primary metrics for performance tracking.

  • Delivery In-Full, On-Time (DIFOT) refers to a number of deliveries performed over the total planned delivery. In addition, the actual deliveries are only considered “actual” if the following four conditions are fulfilled: the delivered product is (1) delivered to the indicated place, (2) according to the quantity ordered, (3) according to the expected quality and finishing, and (4) according to the indicated time.
  • Deliveries per area/time refer to the actual deliveries performed for designated areas or time periods. For deliveries per area, you will have to demarcate areas based several factors, such as customer density and presence of distribution facilities. For deliveries per time, time periods can be determined by fixed time blocks (i.e. daily, weekly, monthly) or seasonal blocks (i.e. December festive holidays). Being able to produce figures for deliveries per area/time allows you to compare the performance of your supply chain across regions and time periods. Thereafter, you can hone in on underperforming areas and attempt to uncover factors contributing to the less-than-stellar results.
  • Weight by customer refers to the total weight of deliveries accrued to a single customer within a defined time period. Consistent gathering of data for this KPI allows you to identify customers who order regularly or order in bulk. Subsequently, you can decide to devote more attention and resources in maintaining these customers who contribute most to growth.
  • Customer order cycle time refers to the time taken to deliver a product to the customer after the purchase order is received. The smaller the customer order cycle time, the more efficient your supply chain as it indicates the speed in which your company can push products out.
  • Distribution of jobs by driver/vehicle refers to the number of delivery jobs assigned to specific drivers or vehicle within a defined time period. There is no industry benchmark, but this KPI allows you to gauge whether the driver or vehicle has been underloaded or overloaded with delivery jobs. Subsequently, you can decide to increase or decrease the number of drivers or vehicle in order to push for an optimal delivery rate and ensure that both your drivers and vehicles are not overworked.

Identify the company’s biggest cost

Now, assuming that a system has been set up to collect data based on the aforementioned KPIs and a few months’ worth of data have been collected, it is imperative that the data is analysed. In particular, factors contributing to slow growth and costs have to be identified in order to drive improvements.

The following indicators are commonly used to identify sources of costs for logistics companies.

Failed deliveries refer to the difference between total planned and actual deliveries. Alternatively, it refers to the difference between the company’s DIFOT rate and the full (100%) of the DIFOT rate, multiplied by the total planned deliveries.

Deliveries are considered “failed” if they are:

  • Not delivered to the specified location
  • Not according to specified quantity
  • Not according to expected quality and finishing
  • Not according to the specified time period.

Failed deliveries are most applicable to the e-commerce sector, which involves multiple orders scheduled for different time periods. Given the tight scheduling of deliveries, the chances of products being delivered outside of specific time periods are relatively high.

For e-commerce deliveries, often, a few delivery attempts must be made before a parcel can be delivered successfully to its recipient. Should your data indicate a relatively high percentage of failed deliveries, in-depth research into the reasons for the failed deliveries must be performed in order to mitigate future reoccurrence.

Inventory turnover refers to the speed in which an item comes in and out of the warehouse. This KPI specifically measures the efficiency of the company’s inventory management system. The higher the ratio, the more efficient your supply chain. However, this condition only holds if you are making a profit from the high turnover. Should you have a low inventory turnover, then there is a need to determine whether the poor performance is due to fixed or variable factors.

Firstly, low inventory turnover may be due to small warehouse capacity, which acts as a bottleneck preventing companies from quickly filling old inventory with the new. Secondly, low inventory turnover may be due to insufficient delivery vehicles or orders to quickly push products out. Lastly, low inventory turnover may be due to insufficient volume of products to replace ones in the warehouse. Identifying and tweaking the factors contributing to inventory turnover are key to agile operations.

Vehicle use rate refers to the ratio between active and idle vehicles. The nearer the ratio is to the value 1, the higher the employment rate of your vehicular and driver resources. Should you have low vehicle use rate against the benchmark that you have set, then a few factors must be considered.

Firstly, there may be insufficient orders to convert into physical deliveries. Secondly, there may be an insufficient number of drivers to deliver the goods, resulting in idle vehicles being underutilised. Thirdly, the aforementioned conditions can be met, but the delivery system may not be optimized, resulting in frequent breaks between deliveries. Nonetheless, vehicle use rate fluctuates and it is imperative to periodically monitor the aforementioned conditions to ensure a consistent high vehicle use rate.

How to Measure?

Choosing the right tools for analysis

Incorporation of data analytics tools is important in making the data analysis more insightful and more efficient. There are multiple software in the market, both paid and free for use, that can assist in revealing data trends and even recommending action to address certain gaps in your operations.

For a start, integrate Google Analytics into the company. Google Analytics is a free web analytics service that tracks and reports on your website traffic. The service contains different types of data visualization that cater to both casual users and companies, allowing a wide area for customization to the company’s needs. Should this be your first time integrating Google Analytics to the company’s operations, there are free Google Analytics lessons and certification tests available to quickly equip the user with the knowledge to use the service and enhance the data analysis process.

An example of a dashboard in Google Analytics. Image Credit: Kissmetrics

For feedback from customers, Google Forms coupled with its integrated Excel function can be used to collect survey responses. This free service also allows for online collaboration, which allows for same-day analysis among team members regardless of locality.

View Google Form responses in Sheets. Image credit: G Suite Learning Center

In order to calculate some of those KPIs mentioned, a Transport Management System (TMS) can be used. Most TMS can track delivery milestones recording the time stamp of when the delivery is completed, which drivers did which orders and the job details, including area and time period distribution. Also, a good TMS will be able to record the reasons for failed deliveries, allowing the company to determine their main cost for failed delivery. Some TMS will also allow the calculation of vehicle utilisation in terms of load optimisation, which can then be used to further optimise the use of the vehicles.

These are a few useful KPIs that can be tracked in a TMS like VersaFleet.
This box plot displays a distribution of deliveries among drivers across two months. A box plot is a highly visual and effective way of viewing a clear summary of one of more sets of data.

A TMS can be further supplemented by a Fleet Management System (FMS), which specialises in vehicle tracking. The consumption of fuel can be monitored. Also, for temperature controlled trucks, the temperature can be monitored on a regular basis. It can also monitor vehicle active and idle rates.

Graph showing fuel level by distance. Image Credit: Cartrack

Nonetheless, usage of analytical tools is insufficient without a proper analytical framework.

One of the staple analytical frameworks used in data analytics is the Gartner’s Model, which shows the value created for a business as the maturity level of its analytical system increases.

Using this model, the goal for companies is to move from “descriptive analytics” to “prescriptive analytics”, or in other words, companies can move from simple identification of problems to the next level of implementing solutions to improve the operations.

Gartner’s Analytics Maturity Model. Image credit: New Signature

Identifying the stage at which the company is at in terms of your data analytics capabilities using the Gartner’s model can help to chart the company’s operational growth roadmap. Thereafter, data analytics will become an indispensable part of the company in driving growth.

Nonetheless, there are different analytical tools for different stakeholders of the company. Exploring and determining the most suitable tool to analyse collected data is imperative to reveal key market trends that the company can effectively capitalise upon.

Set up a digital infrastructure

Apart from determining the operational KPIs and choosing the right tools for data analysis, setting up a strong and adequate digital infrastructure in the company will allow the company to handle a higher bandwidth of data. If the company is just starting out, a single desk workstation and backups of collected data are sufficient to perform basic data analysis. However, once the company starts to grow, a proper digital infrastructure for data analysis is necessary.

At some point, a company can no longer keep all the data on various spreadsheets saved locally in a single computer. When this happens, the best thing to do is to acquire proper tools that is easily accessible through the cloud. Cloud storage is more scalable than a local physical server in the office and would be able to handle the huge amounts of data collected.

In conclusion, dealing with Big Data is not as difficult as conceived by many. Indeed, data analysis involves multiple operational considerations, including identification of stakeholders, generation of key benchmarks, pinpointing of costs, determining the suitable data analytical software and frameworks, as well as the seamless integration of digital infrastructure. However, the rewards that lie at the end of a data-driven growth are immense, making the pursuit of data analytics worthwhile.

 

Do you have a unique take of Big Data in logistics? Let us know in the comments below!