I’ve written many supply chain and logistics research reports in my 27 years as an industry analyst. These reports have covered a wide range of topics, including transportation management systems, supply chain operating networks, third-party logistics, and global trade management. I’ve also read a lot of reports from other industry analysts and consulting firms.
There is a set of learnings and recommendations, typically found at the end of these reports, that have remained constant over the years. Here are the five most common:
Get buy-in from upper management. Without support from senior leadership, which often includes the CEO and CFO, most initiatives will stay stuck at the starting line or won’t get very far due to a lack of investment in people, technology, and other resources critical for success. Executive sponsorship also helps when difficult decisions need to be made, priorities compete, or an initiative crosses functional boundaries.
Don’t underestimate change management. Getting buy-in from employees on the front lines — for example, those who will be expected to use a new software solution or follow a new process — is extremely important. If they don’t understand why the change is necessary or see the value (“What’s in it for me?”), they will hang on to “the way we’ve always done things” for as long as they can. Involving users early in the process, listening to their concerns, and providing adequate training can make the difference between adoption and resistance. Technology implementations ultimately succeed or fail as much because of people and processes as because of the technology itself.
Get the data right. We’ve all heard the expression “garbage in, garbage out,” but it has become even more relevant in the age of AI. Whether you’re implementing a TMS, creating a digital twin, deploying predictive analytics, or training an AI agent, the quality of the output depends heavily on the quality of the data going in. Incomplete, inaccurate, outdated, or fragmented data will undermine even the most sophisticated technology. Data quality and governance, therefore, can’t be treated as somebody else’s problem or as something to clean up later.
Start small. The classic crawl, walk, run. Few companies today have the appetite for the cost and risk associated with “big bang” implementations. Identify a relatively low-risk, easy-to-deploy use case first, prove the value, learn from the experience, and build from there. Early successes also help generate broader support for the initiative, including from senior management and front-line employees. Just don’t confuse “start small” with “think small.” You should still have a broader vision of where you ultimately want to go.
Measure results. “You can’t manage what you can’t measure” is another maxim that has stood the test of time. Identify the business problem you’re trying to solve and the KPIs you want to improve at the beginning — instead of, for example, buying a software solution in search of a problem. What does success look like? Lower costs? Improved service? Greater productivity? Less inventory? Faster decision making? Establishing a baseline and defining the desired outcomes upfront makes it possible to determine whether the initiative is actually delivering value and whether you should continue investing in it.
There are others, of course, but these five points have stood the test of time.
Maybe we should come up with a name or acronym for these five learnings and recommendations, so we wouldn’t have to write them out all the time. At the end of our reports we could just say, “In addition to [acronym], we recommend the following…”
Unfortunately, Executive Buy-In, Change Management, Get the Data Right, Start Small, and Measure Results doesn’t produce a particularly catchy acronym. Maybe we should just call them The Constant Five.
What I find interesting is that these recommendations have survived wave after wave of technology trends — from client/server applications in the late ’90s and early 2000s to generative and agentic AI today. The technologies and terminology keep changing, but the fundamentals of successful transformation haven’t changed nearly as much. You still need leadership support, employee buy-in, good data, a pragmatic approach to implementation, and a clear way to measure results.
What have I learned after 27 years of conducting research in supply chain and logistics? The newest technology rarely eliminates the oldest challenges.
So, the next time you read a research report about AI agents, autonomous supply chains, or whatever comes next, skip ahead to the recommendations. I’m willing to bet The Constant Five will still be there.
