Every generation has its big event that changes everything. Millennials happen to have had the “opportunity” to witness several once-in-a-lifetime advances: the internet rewiring the way we access information, the rise of mobile and social media reworking how we connect and shop, and now AI reshaping what a brainstorming session even looks like.
That front-row seat to history leaves a mark. There’s a whole cohort of people leading companies and sitting in boardrooms today whose expectations were molded by watching technology create markets and change daily life seemingly overnight.
That thought kept surfacing while reading The Abercrombie Age, Myles Ethan Lascity’s study of millennial consumer culture at the turn of the century. The book examines how the late-’90s and early-2000s media landscape sold millennials a story that life would simply keep getting better. What the book doesn’t dig into, and what stuck with me, is how those cultural moments quietly reset what an entire generation expects from innovation itself: that transformation should be sudden, dramatic, and tied to a singular event.
How many times has a new trend surfaced, and the immediate response is to stop, or significantly scale back, everything else in motion, so resources, time, and funding can pour toward preparing for whatever shift this trend will supposedly bring? As Paul Strassmann, former CIO of the Pentagon, Xerox, and General Foods, put it, that kind of funding tends to confront management with “extreme demands for which there was no rationale whatsoever.” The assumption is always the same. That the initiative will scale into real revenue or market share, with no one ever pressure-testing that assumption first. It’s not a new pattern. It’s also not a great strategy.
What Did We Learn from Y2K
Y2K is the clearest case study we have. The world spent an almost unbelievable amount of money bracing for it. Estimates range from roughly $100 billion in the U.S. alone to $300–500 billion globally, according to Senate committee findings and industry analysts cited by Computerworld and Slate. In 1999, cities were filled with a mix of excitement, fear, and skepticism, and, more importantly, mass buy-in that something was going to change, and it would be big. Talk about an incredible survive-or-thrive atmosphere.
Then January 1, 2000 arrived, and nothing happened. The big transformation everyone had braced for never appeared. Researchers who later studied the period, including a Federal Reserve Bank of St. Louis analysis, found little evidence that the spending itself drove any deeper, lasting shift in how businesses actually operated.
Which raises the question: does the absence of the catastrophe mean the whole effort failed?
Not exactly. But it’s close. With Y2K, the money had one use, tied to one narrow assumption, that a catastrophe is coming. When the catastrophe didn’t arrive, there was nothing left standing that hadn’t already served its single purpose. The real test was never the event itself. It’s the gap between the expectations a company built going in and what it was actually prepared to do once reality fell short of that expectation. Agile companies aren’t the ones that react fastest to a headline. They’re the ones that build something durable enough to outlast being wrong about the headline.
“Agile companies aren’t the ones that react fastest to a headline. They’re the ones that build something durable enough to outlast being wrong about the headline.”
Familiar Pattern?
AI is clearly not the last disruptor we’ll live through, and we’re already watching the same pattern repeat. A widely cited 2025 MIT study, “The GenAI Divide: State of AI in Business,” found that despite $30–40 billion in enterprise AI investment, roughly 95% of generative AI pilots produced no measurable financial return. The telling part isn’t just the failure rate. It’s what separated the 5% that worked. Those weren’t the biggest bets on the hype. They were narrower, integrated into real workflows, often built alongside outside partners rather than sized to match the “this changes everything” headline.
Two companies show what that looks like in practice. Morgan Stanley didn’t announce an AI transformation. It worked with OpenAI on one narrow job. To help financial advisors find answers in the firm’s research. It piloted with a few hundred advisors before rolling out to thousands, and the firm reports that more than 98% of advisor teams now use it. Lumen Technologies took a similar route, putting Microsoft’s Copilot inside the sales workflow its sellers already used for customer research, meeting summaries, and CRM updates. The company estimates it saves each seller about four hours a week. Neither bet depended on the hype being right. Both only had to be useful on a Tuesday afternoon. In other words: built to be useful regardless of whether the broader promise ever arrived on schedule.
The Real Test of Agility
So, back to the question: when what you’re planning for doesn’t play out as expected, is the transformation a complete failure? No. Not if the goal is to actually be an innovative company rather than just look like one for a news cycle.
Agile companies aren’t distinguished by how they react to a headline. They’re distinguished by how they prepare to be wrong about it. Did they build something that survives disappointment and still serves the mission? Or did they only build for the one version of the future where the hype turned out to be correct?
This tracks with what change management research has argued for decades. Most organizational change doesn’t fail at the announcement. It fails in the unglamorous stretch afterward, when no one’s watching anymore, and the easiest thing to do is let old habits creep back in. Y2K nailed step one and skipped straight to the end credits.
Worth thinking about before the next headline gets the budget.
Sources:
The Abercrombie Age: Millennial Aspiration and the Promise of Consumer Culture, Myles Ethan Lascity (UNC Press); Computerworld, “Y2K: The good, the bad and the crazy”; Slate, “Was Y2K a Waste?”; Los Angeles Times, “U.S., Firms Overreacted to Y2K Fix, Critics Say” (Jan. 2, 2000), as reported via ACM TechNews; Federal Reserve Bank of St. Louis research on Y2K-era business investment; MIT NANDA, “The GenAI Divide: State of AI in Business 2025”; OpenAI, “Morgan Stanley uses AI evals to shape the future of financial services”; Morgan Stanley press release, “AI @ Morgan Stanley Debrief” launch; Microsoft Source, “How Copilot is helping propel an evolution at Lumen Technologies.”
About the Author:
Briauna Perryman is a Senior Consultant at Kotter.
