The Proof is in the (Data) Pudding: A Look at ASN as an Inaccurate Performance Measure

By: Azad Sadr

In a recent guest post about supplier performance metrics, I raised the point that time-to-ship provides much more accurate performance data than tracking code creation (ASN) timestamps.

I also noted that performance metrics based on tracking code creation data incentivised suppliers to create such codes as soon as possible in order to avoid SLA penalties rather than achieve actual on-time shipment of orders.

In this post I’d like to go through some real-world data to support those assertions.

Instead of analyzing our entire data warehouse, I’ll take a look at the fulfillment and shipping data of roughly 10,000 orders from a recent Tuesday in November.

This will demonstrate that the weaknesses of ASN as a data point for on time shipment performance is apparent even in just a single day’s worth of data.

Histogram

Here’s a histogram of the overall tracking code variance in our data:

 

The most important take-away from the above histogram is that only 36% of tracking codes are created within 5 hours of pickup.

For well over half of the drop shipped orders, therefore, tracking code creation does not come close to indicating the completion of the supplier’s role in fulfillment.

A Deeper Look at the Data:

Roughly 10% of orders were given tracking codes 24 hrs before pick up by carriers.

One supplier even created four tracking codes over 200 hours (that’s more than eight days!) before the orders were shipped. Another supplier created a tracking code one minute after order creation but still took 34 hours to ship the item out.

This leads to three important visibility issues.

1) In terms of SLAs and compliance, there is sometimes such a large variance between tracking code creation and actual time to ship that it becomes difficult to use ASN as an accurate gauge of on-time shipment.

It also makes it easy for suppliers to technically remain in compliance even though an order was shipped late.

2) As for inventory visibility, creating too many tracking codes for too many retailers could cause suppliers to think they have more inventory of a particular SKU than is actually available, raising the risk of cancellations.

This is not an idle concern. In the United States it is estimated that the accuracy of inventory levels that suppliers give to retailers is only around 55-65%.

3) Finally, a customer who has received a tracking code from their retailer will think that their order has shipped even though it may still be waiting for several days in a warehouse to be picked up by a carrier. This could lead to bad customer experience

9% of orders were reported five or more hours after carriers had already picked them up.

One supplier even took 186 hours (again almost eight days!) to return a tracking code for a particular order.

This leads to two issues.

1) First, depending on SLA agreements, suppliers might be penalized by retailers for non-compliance even though their orders have been shipped on time. This is bad for supplier relations.

2) Second, forcing a customer to wait 15 hours before receiving a tracking code can lead to bad customer experience. In some instances they might even receive their package before they’re able to access shipping information.

68% of all upgraded orders occurred within a time-to-ship window of 0-70 hours, while 32% of upgraded orders occurred after 70 hours.

These upgrade statistics get interesting when compared with data that uses tracking code creation as the data point for judging on-time shipment.

Based on tracking code creation, 84% of all upgraded orders occurred within a window of 0-70 hours, while only 16% of upgraded orders occurred after 70 hours. Tracking code creation data therefore skews upgrade data towards the 0-70 hour time period.

This is important because the root cause of upgrades are extremely time-based, with upgrades that occur later in a shipment lifecycle more often the result of inefficient supplier fulfilment processes than those that occur earlier.

Such tracking code creation data could therefore encourage retailers and suppliers to focus much fewer resources on preventing late stage upgrades even though they still account for over 30% of all upgrades.

That translates directly into higher overhead.

Preventing just 30 late stage upgrades per day could save a retailer over $200,000 in shipping costs.

ASN Is a Measure of Process, Not Results

Based on the above data points, it’s clear that tracking code creation is not an ideal data point for judging on time shipment performance.

The reason for this is that tracking code creation is not really a measure of results but rather a measure of process, i.e. the suppliers tracking code creation process.

The manner in which a supplier fulfills orders, however, should not matter to retailers as long as they ship out orders on time.

What is needed instead is time-to-ship data that actually measures the performance results that retailers are looking for: on-time shipment.

But if that’s the case, then why aren’t more retailers using time-to-ship in their SLAs?

For many retailers, the biggest obstacle to using this more accurate data is getting access to it.

Even when shipping data is available through carrier or third party connections, correlating such external data with their own order data requires IT resources that many retailers are unable to allocate away from other higher priority projects.

In the coming year we’ll be adding some cool data features to the Dsco platform that will allow trading partners to not only access such data but perform just these types of correlations with their own data in real time to tackle issues such as excessive upgrades, late shipments, and cancellations.

In my next post, I’ll use some of these new data capabilities to highlight aggregated fulfilment and shipping patterns that can allow retailers to provide more accurate time windows to end customers as well as prevent a variety of fulfillment and delivery issues before they occur.

Stay tuned.

Azad Sadr

Azad Sadr heads up Dsco’s Content and Product Marketing departments. He loves telling stories and boiling complex topics down to digestible content for non-specialist readers. Over the course of his career he’s worked in academia, the oil and gas industry, and most recently, supply chain research and analysis. In his free time he enjoys walking around the town, drinking coffee, and writing fiction.