Today you are going to learn which analytics and metrics matter for ecommerce supply chains.
This resource is different from anything else you’ll find because our goal is to list and talk through lots of examples. Few resources out there extend beyond a bland definition of analytics.
The guide will be especially interesting to ecommerce professionals because the list is sourced from veterans of Amazon and Zulily, pulling from their experience of measuring the things that mattered.
There are a lot of ways to measure the performance of your supply chain. Some are better than others. What’s most important is the understanding that an ecommerce chain is dramatically different from the ones that support older, physical business models, like brick-and-mortar (B&M) stores.
Let’s briefly talk about how to think about supply chain analytics (and metrics, KPIs, and measurement) in an ecommerce sense; then we will dive into the long list of possible options for you to pick for your own company.
Key highlights:
Supply chain analytics refers to the process of collecting, analyzing, and interpreting data across the supply chain to improve efficiency and enhance leaders decision-making process.
This process involves descriptive analytics (what happened), diagnostic analytics (why it happened), predictive analytics (what will happen), and prescriptive analytics (what should happen).
Veterans of large ecommerce companies intimately understand what separates supply chain data analytics from metrics, KPIs, and measurement, but let’s quickly give a definition so the following list is easier to understand.
The definitions build on each other, almost like a pyramid.
Measurement is the “how” of data collection. Measurement within ecommerce businesses is uniquely different — and better — than in physical stores. Since ecommerce systems are inherently digital, data is being shed with every “action” happening along the supply chain. Therefore, measurement is as simple as capturing the enormous amounts of data produced from every interaction within every system every second of every day.
Some examples:
Get the idea? The measurement of all those actions is how modern ecommerce companies gain sophisticated metrics and analytics that set them apart from the competition. It’s really hard for Office Depot to know which pen the customer wanted when passing through Aisle 12 at the store down the street, for example.
The big “ah-ha!” moment for ecommerce companies is data = measurement, and ecommerce companies inherently have a lot of data! To take advantage of it, and create metrics and analytics that boost your business, you need to store the data in a structured way and connect it with other sources of data.
Metrics are quantifiable measurements that give a standardized value. For example, “Out-of-Stock-Bounce-Rate” is a metric, and is the culmination of the measurements from our previous story. By stringing together the data, a supply chain manager can create this metric with a statement like “17% of the time a customer visited a product page, we were out of stock, and they immediately left the site.” That’s a metric.
Key Performance Indicators (KPIs) are simply your most important metrics. It might be stock levels, inbound cost-per-unit, NPS scores, whatever. You might check other metrics monthly, but you look at KPIs on a weekly or even daily basis.
Analytics adds context to a collection of metrics to inform future decision making. Continuing our story, let’s suppose the supply chain manager ended contracts with 10 suppliers last month because the company didn’t believe there was demand anymore for those 10 products. But by digging into the “Out-of-Stock-Bounce-Rate” and combining it with other data, the manager notices that over half of the bounces are for one of the suspended products. The manager can take this insight, demonstrate a missed revenue opportunity, and perhaps restart negotiations with that supplier.
Read our complete ecommerce supply chain guide.
There are four types of metrics and analytics, but for the purpose of this next section, we’ll just call everything analytics.
Each type builds on the previous type towards more automated decision-making. While full automation is a dream scenario for many businesses, it simply isn’t practical, so all four types play a role in a modern data-driven company.
Descriptive analytics explain “what happened” at some point along the supply chain. Examples:
Descriptive supply chain analytics are the bulk of what makes up dashboards. While all four families are useful for tracking month-over-month changes, descriptive analytics are especially useful because of the guaranteed consistency of the measurements.
These supply chain data analytics are inherently simple measurements with little interpretation, which means they are less useful at automating decision making and more useful at logging rates-of-change for basic measurements over a given time period. For example, an increase in weekly inventory spend may alert you to a change happening that your partners elsewhere in the business may have forgotten to tell you about. Descriptive analytics are the foundation of all other analytics and are frequently critical in their own right.
Diagnostic analytics explain why something happened. Examples:
Diagnostic analytics are most often used to assess root causes of past output or outcomes. They tend to be more of a tool you use on a specific problem versus a regular measurement over time, like descriptive analytics. You will see them deployed to understand what went wrong if a particular outcome was unexpected.
Diagnostic analytics are usually a combination of several descriptive analytics to tell a story about causation.
One of the more interesting areas we see diagnostics analytics is within simulation exercises. Few people have the bandwidth to do these exercises on their own, but they can be useful when available. A specific scenario can be played out via a simulation in order to create the upside of diagnostic supply chain data without the downside of a real-life negative outcome. Simulating a pandemic to diagnose the impact on customer demand or supplier lead time sure would have been useful to do in January 2020!
Supply chain predictive analytics explain what will happen in the future. Examples:
The most important thing for predictions, whether by human or machine, is having lots of data, as stated by the MIT Sloan Management Review. Predictive models need historical data to project what will happen in the future. Companies that have high quality data sources that are interoperable and structured will perform better at predictive analytics.
While segmentation can happen at any layer, it has the biggest impact on company operations with predictive insights. It’s common to segment predictive analytics by product, customer, channel, geography, and so on. This is the kind of analytics that will tell you not to purchase winter coats for Miami during summer, for example.
Prescriptive analytics tell you what actions to take with some degree of confidence. Examples:
Prescriptive analytics are probabilistic. Much like predictive analytics, they take a ton of data and then model outcomes. But they also consider probabilistic scenarios and suggest the best actions to take (this is what makes them prescriptive).
These supply chain metrics are deployed mostly related to performance outcomes. Minimizing shipping costs, increasing revenue and reducing in-stock levels depending on the situation are all good examples of use cases for prescriptive analytics.
Everyone wants prescriptive analytics because of their power to automate decision-making, but we are still in an era where the practicality of that is less than marketers would have you believe. Human consideration and decision-making are still required for many tasks because of how complex supply chains and logistics can be. But prescriptive analytics go through the brute computational work for you to help make decisions easier and more informed than may be possible through other means.
Instead of jumping right to perspective analytics, you can gain immediate value by getting accurate descriptive and diagnostic analytics in place first, without which prescriptive analytics are not possible.
Based on the linear flow of goods, we’ve listed 40 example analytics and metrics to help you paint a broad picture of what’s possible — and advised — for ecommerce companies.
Main inbound supply chain metrics include:
And so on. There is no one best practice or universal goal for this metric, but instead targets are frequently determined by geography. For example, Japan prioritizes more frequent orders at lower volumes and higher costs-per-unit due to the cost of real estate and overall storage costs-per-unit, whereas American companies receiving products from China will almost always benefit from waiting to fill an entire container before shipping across the Pacific Ocean.
We generally define the supply chain as the flow of goods, and in the context of ecommerce, often a single-piece flow. Modernizing the supply chain means optimizing the continual movement of goods.
Inventory management, much like demand forecasting, is its own science. Indeed, there are whole books written on different ways to measure and manage inventory. The practice is intimately related to the supply chain, but is separate and large enough of a topic that it’s worth saving this discussion for a different article.
Instead of listing inventory management metrics, I’m going to provide a few concepts linking inventory to other metrics throughout the supply chain so that you understand in greater detail how much modern ecommerce supply chain analytics tools rely on the interoperability of data across multiple systems.
Here are key fulfillment metrics logistics leaders should be tracking:
Discover 10 Shopify fulfillment best practices for ecommerce businesses.
Relevant fulfillment metrics in supply chain analytics software include:
Learn how to reduce outbound shipping costs.
For ecommerce companies, returns can be a hard puzzle. Returns are part of modern ecommerce and can’t be entirely avoided. You can improve processes and data to reduce some of them, but you can’t get rid of them, so it’s important that you are good at them in a way that doesn’t completely destroy your margins.
Subsequently, supply chain managers should keep a close tabs on metrics to mature the return process and incorporate them into reverse logistics analytics.
Winning ecommerce companies prioritize metrics that put the customer first. It’s hard to do, though, because they are inherently cross-functional. It requires a shift in company perspective. The result, however, is supply chain analytics that propel the business towards profit and gross margin growth that beats competitors.
Ecommerce logistics move faster than traditional supply chains, requiring agile, data-driven decisions. Shipium is a modern logistics platform designed to help enterprise retailers and ecommerce brands make smarter, data-driven shipping decisions.
Our system ensures retailers shift from descriptive analytics (what happened) to prescriptive analytics (what should happen), enabling them to proactively improve fulfillment speed, reduce costs, and enhance customer satisfaction.
With the Shipium platform, ecommerce brands can move beyond fragmented, outdated systems and centralize supply chain analytics, metrics, and automation in one place. Request a demo today to see how Shipium can elevate your supply chain analytics.
Big data and analytics give ecommerce companies greater insight into their logistical operations, helping to transform conventional supply chain management paradigms. Key changes include:
Modern supply chain analytics tools, like Shipium, integrate real-time data with automation, allowing businesses to shift from reactive decision-making to enhance supply chain performance.
To effectively use analytics in supply chain management, businesses should:
By leveraging data analysis, companies can make smarter, faster, and more cost-effective logistics decisions.
Choosing the right supply chain analytics tool comes down to understanding what your business needs and how data can help you improve operations. Start by identifying your biggest challenges — are you struggling with high shipping costs, inaccurate demand forecasting, or slow fulfillment?
The right tool should give you real-time visibility, help you forecast demand, and even automate decisions like selecting the best carrier or fulfillment center. Look for features like:
Before committing, compare pricing and test-drive a few platforms. Many providers, like Shipium, offer demos, so take advantage of that to see how well a tool fits into your workflow.