When Companies Lose the Plot

There is a particular kind of corporate decline that doesn’t announce itself. No scandal. No collapse. Just a slow drift away from what made the company worth building in the first place, replaced by a relentless chase after someone else’s financial profile.

FedEx just gave us a textbook preview.

This week, FedEx announced plans to build an “AI agent workforce” – bots embedded across operations, marketing, network planning, and software development. 300,000 employees trained. Over half of core workflows touched by AI agents by 2028. A whole governance hierarchy: manager agents, audit agents, worker agents. The CIO called it a “global shift.” The CEO called AI a “true force multiplier.”

It sounds visionary. It isn’t.

What FedEx described is not a strategy. It is an acceleration plan. And acceleration without direction is just a faster way to end up somewhere you didn’t want to be.

I watched this dynamic play out firsthand at a major American telecom. The company spent years in open envy of tech company valuations. The price-to-earnings ratios were intoxicating. So leadership did what companies do when they want to become something they’re not: they acquired, restructured, and optimized. They chased margin. They talked about transformation constantly while making decisions that had nothing to do with serving customers or building durable operational capability.

The culture eroded. The acquisitions made no sense to the people doing the actual work. Leadership became obsessed with the stock price. And through all of it, the company got faster – faster at executing a strategy that was never clearly defined, faster at scaling dysfunction, faster at losing the institutional knowledge and human judgment that had taken decades to accumulate.

AI agents would have made it worse. Not better. Faster. There is a difference.

Here is the core problem with what FedEx announced: agents inherit the objectives you give them. They don’t question the objectives. They don’t reframe the problem. They don’t walk into the planning meeting and say “I think we’re optimizing the wrong thing.” They execute. At scale. With consistency. And if what you’re executing is misaligned with what actually creates value, you will now be misaligned with extraordinary efficiency.

The FedEx CIO said something that should have been the whole presentation: “If I give you bad information, you will make bad decisions.”

Correct. And bad strategy is bad information.

The question every logistics company should be asking right now is not “where can we embed AI agents?” The question is “which decisions in our operation are currently made badly, and why?” That is a problem framing question. It requires honest diagnosis before solutioning. It requires someone willing to slow down long enough to define the right objective before handing it to a system designed to pursue objectives relentlessly.

That kind of thinking doesn’t get announced at investor days. It doesn’t generate headlines. It doesn’t make a CEO sound like they’re riding the AI wave. But it is the only kind of thinking that produces durable results.

Gartner estimates that over 40% of AI agent projects will be canceled by end of 2027 due to escalating costs, unclear business value, or inadequate risk controls. FedEx is a well-run company with real infrastructure and real data. They may beat those odds on execution.

But execution is not the risk here.

The risk is that FedEx, like the telecom before it, becomes very good at moving fast in the wrong direction. That they trade the soul of a logistics company – the operational excellence, the human judgment built into a global network over decades – for a valuation story that was never theirs to tell.

This is what early decline looks like. Not dramatic. Not sudden. Just a company with a great operation deciding that what it really wants to be is a tech company. Pointing its best new tools at that ambition. And calling it transformation.

The agents will be fast.

The question is whether anyone bothered to ask where they’re going.

If this resonated, follow me on LinkedIn where I write about Decision Systems Management, why most transformation announcements are just acceleration plans in disguise, and what it actually looks like to build systems that work. Give it some claps if it was worth your time – and if you want more, subscribe to the Bit Bros newsletter.