Why “Good Enough” Keeps
Producing Predictable Failures
Two phrases dominate leadership conversations right now. I hear them most often when: “We’re directionally correct” or “That was an own goal.”
Both are modern shorthand. Both can be useful. And both are often misused in ways that quietly compound risk inside companies.
Being directionally correct is sometimes the right posture—especially in complex, uncertain environments. I’m a fan of moving quickly with imperfect information and correcting along the way. But when leaders treat the directional movement as an endpoint rather than a waypoint, they unintentionally enable the very outcomes they later regret.
Across companies, most failures are not caused by bad intent or lack of intelligence. They stem from weak execution discipline, delayed correction, and systems that allow small errors to grow unchecked.
There are times when good enough is exactly right.
Signal to Watch: “We’re Too Far In”
If leadership justifies continuing an initiative primarily by referencing sunk cost—time spent, money invested, political capital used—strategy has given way to defense.
Clear Intent, Incomplete Information
Effective companies act with partial data, but they do so transparently. Assumptions are explicit, not implied, and plans are revisited as conditions change. Directionally correct works only when reassessment is built in.
Early Movement with Built-In Review
High-performing companies move early, but they pause deliberately. They establish review points to evaluate what’s working, what isn’t, and what needs adjustment before scale sets in. Progress is allowed. Blind momentum is not.
Reversible Decisions
Strong companies design early decisions to be reversible. Leaders preserve flexibility and avoid turning early judgment calls into reputational traps. “Burning the boats” is only done once.
Named Accountability
Someone owns the call. Someone owns the outcome. “Directionally correct” never means “no one’s responsible.”
When “Directionally Correct” Becomes a Leadership Liability
This is where patterns repeat—and where own goals are born.
Strategy Without Stop Conditions
Many companies launch initiatives with ambition but no defined criteria for pause, reset, or termination. Problems are acknowledged, but action is deferred because “the direction is still right.” Singer Kenny Rogers said, “…know when to fold ’em.” Professional poker player Annie Duke said it with the one-word book title, “Quit.” Professionals employ stop-losses.
Escalation of Commitment
Once leaders invest time, money, or reputation in a course of action, they become less willing to challenge it—even as evidence mounts. Perseverance quietly turns into denial. When I’m brought into situations like this, the solutions appear obvious and immediate. It’s the psychology of framing it for the executive to see that taps into my expertise. That’s culture masquerading as weak employees.
Suppressed Signal
In underperforming companies, the people closest to the risk often see the issue first—but don’t feel safe escalating it. Companies don’t fail because people don’t know. They fail because people don’t speak.
Process Over Judgement
Meetings continue. Reports multiply. Dashboards stay green. Meanwhile, the underlying logic erodes. By the time leaders admit the issue, the cost is no longer marginal—it’s structural.
What an Own Goal
Actually Looks Like in Companies
Own goals are rarely dramatic at the start.
They are self-inflicted and cumulative.
Solving the Wrong Problem Very Well
Teams execute flawlessly against a misframed objective. Everyone is busy. No one is right.
Incentives That Undermine Strategy
Speed is rewarded when accuracy is required. Compliance is rewarded when judgment is needed. We often think of incentives influencing compensation; what else are you quietly incentivizing with your language and processes?
Silencing the People Closest to the Risk
Early warnings are dismissed as noise or negativity. By the time leadership notices, the damage is already structural. What’s the built in bypass that allows a problem to surface outside the org chart?
Confusing Motion with Momentum
Deliverables multiply. Progress stalls. Activity is mistaken for advancement. Meeting counts in Washington are a fool’s metric.
Delegating Without Guardrails
Authority is pushed down without clarity, context, or constraints. Teams score against their own interests—often proudly. Just Do It worked for Michael Jordan and Nike but isn’t the mantra you want to instill.
How Strong Companies Prevent Own Goals
Companies that avoid self-inflicted damage do a few things consistently.
Define Kill Criteria Up Front
Before launching, they ask: What would cause us to stop or change course? If you can’t answer that, you’re not managing risk—you’re hoping.
Institutionalize Dissent
Not as a personality trait, but as a process. If no one is required to challenge assumptions, no one reliably will. This doesn’t mean assembling a Red Team for every decision.
Audit Outcomes, Not Intent
Good intentions don’t negate bad results. Serious companies review what happened, why, and what to change—without theatrics. The process of the thoughtful debrief makes everyone better.
Make Course Correction Visible
Effective leaders change direction publicly and calmly. This signals competence, not weakness. “Smart people change their minds” is another way to say this.
Signal to Watch: Silence in the Room
If the same voices dominate while others stay quiet, the company is not aligned—it’s cautious. Silence is rarely agreement. More often, it’s risk going unreported.
10 Questions Leaders Should Ask Their Leadership Teams:
- What would cause us to stop or materially change this initiative?
- What assumption are we most exposed on right now?
- What evidence would tell us we’re wrong?
- Who benefits if this continues—and who bears the cost if it fails?
- What are we optimizing for that may be undermining something else?
- Where are we mistaking activity for progress?
- Who is closest to the risk—and have we heard from them directly?
- What’s the smallest test that would meaningfully reduce uncertainty?
- What decision have we made harder to reverse than it needs to be?
- If this becomes an own goal, what warning sign will we wish we’d acted on sooner?
Being directionally correct is often necessary.
Allowing own goals is always optional.
Most companies don’t fail from lack of ambition. They fail from tolerating avoidable, self-inflicted damage.
What Am I Reading?
Empire of AI, by Karen Hao, is one of the most clear-eyed and consequential examinations of artificial intelligence to date—not because it is alarmist, but because it is disciplined. Hao does not speculate about distant futures; she documents how we got here, who made the key decisions along the way, and why those decisions now shape power at a scale most societies are unprepared to govern.
What stands out immediately is the book’s command of facts. Hao brings journalistic rigor to a domain often dominated by hype, corporate myth-making, or abstract ethics debates. She traces the evolution of AI through funding choices, institutional incentives, regulatory gaps, and the quiet consolidation of authority among a small group of actors—individuals and firms who set norms early, faced little constraint, and now operate with enormous leverage over markets, labor, information, and even state capacity.
The most unsettling insight in Empire of AI is not that AI is powerful—it’s that its trajectory was neither inevitable nor democratically shaped. Hao shows how decisions made in a permissive regulatory environment allowed private interests to define the rules of the game before public institutions fully understood what was at stake. The result is a concentration of influence that mirrors historical empires: asymmetrical power, weak accountability, and narratives that frame dominance as progress.
For readers who care about governance, national security, markets, or the long-term health of democratic systems, Empire of AI is essential reading. It doesn’t tell you what to think—but it makes clear that pretending this moment is neutral, technical, or apolitical is no longer an option.
