What You’ll Find This Week

Every organization I've ever worked with on innovation initiatives has a version of the same story: they ran an innovation initiative, it didn't quite work, and then they moved on.

Or at least, they told themselves they moved on.

The unfortunate truth is that even though the initiative ended, the cost keeps showing up. Every failed experiment leaves a residue: in the room where the next proposal gets pitched, in the leader who learned to say no faster, in the person who was the loudest champion last time and isn't raising their hand this time. Nobody tracks it. Nobody accounts for it. It just compounds quietly in the background.

In engineering, that's called technical debt. Innovation carries a similar debt, and I think it's one of the biggest untracked costs in most large organizations today.

Here’s what you’ll find:

Don’t Miss Our Latest Podcast

This Week’s Article

Software teams have a word for the cost of past shortcuts: technical debt. Every workaround, every patch, every decision made in favor of speed today or deferring complexity to "later" compounds quietly until it can't be ignored. The concept gave engineers something useful: a shared language for invisible drag. Something they can identify, prioritize, and argue for resources to fix.

Similarly, every shelved pilot leaves something behind. Every transformation program that ran out of air, every innovation lab quietly folded back into the business, every bold initiative that ended with a lessons-learned slide nobody read. The residue doesn't disappear. It shapes how the next idea gets received, how much runway it gets, how many people are willing to put their names on it. And it accumulates, compounding over time.

But unlike technical debt, nobody's tracking Innovation Debt.

Technical debt accumulates because systems don't self-report the cost of past shortcuts. Like technical debt, Innovation Debt is the residual drag on future efforts that results from past decisions, experiments, and failures. And it accumulates the same way: the cost of a failed initiative disperses across the organization invisibly, settling into culture, risk appetite, and institutional memory.

No balance sheet captures it. No one is accountable for it. It just accrues.

Words like transformation, innovation, and disruption don't arrive neutral inside most large organizations. They carry the residue of every past initiative that used them and failed. So people learn to work around them. New ideas get pitched in deliberately vague terms, described as "efficiencies" or "process improvements" to avoid triggering the wrong reaction. Ambiguity becomes a survival strategy. The org loses its ability to talk clearly about what it's actually trying to do.

A CFO who watched $4M evaporate on an innovation lab develops a pattern match. It doesn't matter that the next proposal is different, better-scoped, or more grounded. The match triggers anyway. The veto is reflexive, not analytical. The more senior the person, the more potent the antibody.

And it compounds: the first big bet got 18 months and real resources. The next got 9. The one after that got a quarter and a "show us traction before we commit." Each failure moves the bar higher and the timeline closer, regardless of the quality of what's being proposed.

Large organizations don't promote visionaries. They promote people who have demonstrated they can protect and optimize what already exists. This isn't a flaw in the system, it's the system working as designed. The problem is that the same instinct that makes someone excellent at running a machine makes them poorly suited to questioning whether the machine is doing the right job.

Steve Jobs built Apple into the most innovative company in the world. Tim Cook succeeded him by being an exceptional operator, and he built Apple into the most valuable company in the world. John Ternus has spent 25 years at Apple making existing products better. But, notably, he has no new product categories to his name. Neither choice is wrong. But when the most innovation-associated company in the world selects its next CEO and the defining qualification is execution on what already exists, that tells you something about what large organizations select for. Even the best ones.

The people who do push innovation inside large organizations are usually doing it against institutional resistance, with resources that never quite materialize. When it fails, they regroup and try again. But each attempt leaves a scar. At some point the cumulative pain and exhaustion win out. And when that happens, the organization doesn't lose a project, it loses the person who was willing to push ahead in spite of all odds.

Most large organizations pushing AI today made an attempt at AI three to five years ago. The chatbot that got launched and quietly retired. The GPT-3 pilot that produced a demo nobody used. The press release announcing an AI strategy that was never really a strategy.

IBM spent $5 billion acquiring healthcare companies to build Watson Health into an AI-powered diagnostics platform. They sold the whole thing in 2022 for $1 billion. The technology didn't work, the market didn't follow, and IBM's reputation as an AI leader went with it. Now they're pushing watsonx as their enterprise AI platform: same company, same brand, new bet, unresolved debt.

McDonald's ran an AI-powered drive-thru ordering pilot across more than 100 locations. The system couldn't handle accents, mixed up orders from adjacent lanes, and ignored corrections. A franchisee summarized it plainly: the technology "will have to be at least 95% accurate and will have to save money over having a person in the drive through, and the way it is designed now, does neither." McDonald's pulled the plug in 2024. The pilot is now a permanent exhibit at the Museum of Failure. McDonald's is still pursuing AI ordering through other vendors. Same problem, new attempt, unresolved debt.

Those efforts didn't just cost budget. They generated debt. And that debt is still on the books.

The competitive pressure to innovate has never been higher; AI is just the latest place that pressure is rearing its head. And most organizations are meeting it the same way they've met every prior wave: with poisoned language, gun-shy leadership, and a shrinking pool of people still willing to push.

They're starting a new initiative on top of unresolved debt. The ideas might be good. The strategy might be sound. But the organization running it is slower, more skeptical, and more risk-averse than it was the last time. And nobody's accounting for why.

Technical debt has a payment mechanism. You staff a team, you allocate sprints, you fix the foundation. It's slow and unglamorous, but it's legible. You can point to it on a roadmap and argue for the resources to address it.

Innovation debt doesn't work that way. There's no refactoring sprint for institutional memory. No ticket you can file for engineering to dissolve leadership antibodies. No tooling that fixes the fact that the word "transformation" now makes half the room flinch. The debt is cultural, and culture doesn't get changed quickly, at least not without upending large portions of the organization itself.

That's what makes the interest rate so high. Technical debt slows your engineers. Innovation debt slows your entire organization's ability to attempt anything new. And unlike technical debt, which is at least visible to the people carrying it, innovation debt is largely invisible. The CFO who vetoes the next proposal doesn't think of themselves as a symptom of past failure. They think they're being prudent. The manager who stops raising their hand doesn't connect that decision to the initiative that burned them three years ago. They've just learned how things work around here.

The compounding effect is real.

Each failed initiative makes the next one harder to fund, harder to staff, and harder to sustain. The organizations that feel this most acutely are usually the ones that went big early, announced loudly, and had to walk it back publicly. The ones that feel it least are the ones that never generated much debt to begin with, because they kept the bets small enough that failure didn't echo.

It doesn't have to work that way. Mike Todasco ran innovation at PayPal for five years out of a backpack. No desk, no formal mandate, stolen conference room furniture. When experiments didn't work, there was nothing to post-mortem and no one to blame. He built failure expectation into the model from the start. Most things won't pan out, so structure things so that most things not panning out doesn't cost you the next attempt. It's not a strategy for every organization. But the principle is: the smaller the bet, the smaller the debt when it fails.

Most large organizations do the opposite. They treat innovation like a capital project. Big budget, formal mandate, executive sponsor, press release. And then wonder why, after a few cycles of that, the org has stopped believing.

Innovation debt doesn't announce itself. It shows up as a meeting where a good idea quietly dies. A proposal that never gets written because the person who would have written it already knows how it ends. A leadership team that talks about innovation constantly and moves on it slowly, without understanding why.

The first step isn't a new initiative or a restructured team or a revised budget. It's recognition. If you can see the debt, you can at least stop adding to it. If you can't, you'll keep running the same cycle: big bet, public failure, compounding drag, repeat.

Most organizations are somewhere in that cycle right now. The ones that aren't are the ones that learned to see it.