“AI” was the buzzword of 2023 and 2024 — and for good reason.

AI’s potential to improve how parcel shippers work is undeniable. Most shippers already rely on a number of software tools (transportation management systems, inventory management systems, audit solutions, logistics intelligence, etc.) to make their lives easier.

While these tools help shippers do work more efficiently, next-generation tools powered by AI will do more work for shippers, allowing them to focus on other responsibilities.

Here’s what you need to know to make the most of this new generation of parcel tech.

What Kind of AI Are We Talking About?

When you hear “AI,” there’s a good chance you picture a conversational large-language model like ChatGPT, or perhaps an image-generator like Adobe’s Firefly.

Those newer tools employ “Generative AI,” which generates new content based on prompts. They have exploded onto the scene recently, but follow their longstanding counterparts, “Traditional AI.”

Traditional (AKA Narrow) AI completes tasks based on predetermined algorithms. These have been around for years (think of a chess-playing algorithm as an example). A specific subset of this category is Predictive AI, which uses historical data to make predictions.

While Generative AI (particularly large language models) is rapidly advancing in the parcel technology space, Traditional AI is making a big impact today.

Traditional AI in Parcel Tech

Parcel software companies use a number of AI models to power their solutions. Here are a few popular ones, and their specific use-cases in parcel shipping.

Nearest Neighbor:

• What it is: A model that uses data proximity to identify similarities and make predictions. Social media sites use this AI model when showing you content similar to posts you’ve interacted with before.

• Parcel use case: Nearest Neighbor can show where you fit in the parcel landscape. One example is measuring your average transit times compared to other shippers in your industry, which can uncover competitive advantages/disadvantages.

Random Forest:

• What it is: A prediction model that uses decision trees to come to one conclusion. Banks use this to assess a customer’s risk.
• Parcel use case: Logistics Intelligence tools use Random Forest for a number of predictive models. One is distribution network optimization, which can project the shipping cost impact of a new send-from location, ensuring a shipper chooses the best city for a new DC or 3PL partner.

K-Means Clustering:

• What it is: A model that groups similar data points to identify underlying patterns and make predictions based on that. An example you’re probably familiar with is Netflix’s “users who watched this also watched” recommendations.

• Parcel use case: K-Means Clustering allows shippers to compare themselves to companies with truly similar shipping profiles (not just similar spending) and determine what “best-in-class” metrics really look like.

Finding AI Tools

Now that we’ve covered the basics, what should you look for in a software partner that uses these AI models?

The most important rule - don’t share your data with just any tool. Search for parcel-specific solutions for security purposes and more relevant results.

AI models, at their core, use a sophisticated framework to give structure and understanding to the underlying data. If not using an industry-specific framework, and not trained on industry-wide information, the model provides limited value.

“Industry-wide” is important here. Many businesses will opt to build out their own AI models to monitor their parcel operations, but if it’s an internal-only tool, it will lack the greater context needed to get the best results.

A company’s shipping profile is a single point on a map. Third-party parcel AI tools use industry-wide inputs to fill in that map and can tell a single shipper where they fall. This helps them know how their performance compares to peers, where they need to improve, and set goals to make the biggest impact. Without this context, parcel KPIs are just numbers.

Another advantage for parcel-specific software is data security and privacy. Sharing data from carrier invoices with publicly available tools (like a ChatGPT) gives the benefit of learning from other data sets, but also potentially exposes your data to others.

On the other hand, industry-specific software providers understand the intricacies of parcel invoices, and the sensitivity of the data within them. While user data improves the accuracy and acumen of their AI models, specifics are never shared with anyone else.

AI Is the Future

Parcel software powered by AI is quickly matching - and even enhancing - what a traditional consultant can do.

The parcel world is more complex than ever. To take action in a timely manner, sophisticated software to analyze that complex data is becoming a requirement.

Some might question if software and AI can match the human expertise of someone with years of experience in parcel shipping. When that AI tool was designed by those very parcel experts, and informed by inputs from thousands of shippers, it certainly can.

When an AI-powered platform puts the knowledge of a consultant at a shipper’s fingertips, they can do that work themselves, allowing a process of never-ending monitoring and optimization.

The future isn’t just about efficiency — it’s about evolving with smarter tools that drive continuous improvement. Embrace AI to stay competitive and unlock new opportunities in parcel shipping.

___

Cameron Clark is the manager of data science & analytics at Sifted, where he builds complex systems that bring power to the potential of shipping data. By creating tools that transform raw information into powerful insights, Cameron’s team ensures SiftedAI can help shippers continuously push for 100% parcel optimization.

This article originally appeared in the January/February, 2025 issue of PARCEL.