What Happens When You Answer the Wrong Question Well?

Great Frame

Competence can hide a mistake better than confusion can.

Framing the Question

Answering the wrong question well is dangerous because it does not look like failure. It looks like progress: a polished plan, a cleaner dashboard, a faster process, a persuasive presentation. The problem is that excellence aimed at the wrong target can move people farther from what matters while making the mistake harder to notice. A poor answer invites correction; a brilliant answer can earn funding, praise, and repetition.

When Competence Becomes Camouflage

What happens when you answer the wrong question well? You become efficiently wrong. You may solve a measurable problem while worsening the real one, and because the result is coherent and impressive, people are more likely to trust it.

A weak answer usually meets resistance. The spreadsheet has gaps. The argument feels thin. The prototype fails in testing. But a strong answer can suppress the very doubt that would have corrected the question. It gives an assumption a budget, a deadline, a team, and a success metric.

That is the trap: the quality of an answer and the worth of the question are separate judgments. A flawless route is no comfort when the destination was mistaken.

The Question Chooses What Counts as Success

Every question silently defines a scoreboard. “How do we increase clicks?” rewards clicks. “How do we reduce call time?” rewards short calls. “How do we get this proposal approved?” rewards approval. None of those aims is automatically wrong. They become wrong when the proxy replaces the purpose.

This is why the deepest errors in work are often not errors of effort. They are errors of target selection. People can be diligent, talented, ethical in ordinary ways, and still be pulled into making the wrong number rise.

The counterintuitive insight is that competence increases the risk once the question is faulty. An amateur may abandon a bad objective after producing poor results. A highly capable team can keep improving its machinery, generating increasingly persuasive proof that the machinery is working.

Wells Fargo: Winning the Metric, Losing the Purpose

Wells Fargo offers a sharp example. In its 2016 enforcement action, the Consumer Financial Protection Bureau said that, spurred by sales targets and compensation incentives, employees boosted sales figures by opening unauthorized deposit and credit-card accounts. The bank’s own analysis found more than two million accounts that may not have been authorized by consumers; the CFPB imposed a $100 million penalty.

The implicit operating question was not written on a wall, but the incentive system effectively asked: “How can we produce more account openings?” Employees produced an answer that made the sales metric move. The customer-centered question — “Are we helping customers choose and use products they knowingly want?” — was not governing the scoreboard. This is an inference from the sales incentives and consumer harm described by the CFPB.

That distinction matters. This was not simply a case of people failing to execute. It was a case in which a metric could be successfully increased while trust, consent, and customer welfare were violated. Answering the narrower question well created evidence of performance until outside scrutiny exposed what the performance meant.

The Everyday Version: A Support Dashboard That Lies

Imagine a software company with twelve support agents and a painful backlog. On Monday, the operations manager asks, “How can we cut average ticket-closing time below eight hours?” The team introduces response templates and rewards quick closures. By Friday, the dashboard turns green: closing time falls from thirty-six hours to six.

Two weeks later, renewal calls reveal a different story. Customers whose single-sign-on setup failed received quick instructions that did not fit their configuration. They reopened tickets twice, missed onboarding deadlines, and described support as “fast but useless.”

The team answered its question well. The wrongness lived in what the question omitted: resolved customer work, repeat contacts, and confidence. A better operational question would have been: “How can we restore a customer’s ability to complete onboarding quickly, without creating repeat help?”

The Target Integrity Test

Before polishing an answer, test whether the question deserves optimization. Use four checks:

1. Purpose: What real-world outcome is this question standing in for? If “tickets closed” stands in for “customers unblocked,” write both down.
2. Gaming: What behavior would improve the metric while betraying the purpose? Name an unethical version and a merely shortsighted version.
3. Missing cost: Who could pay for our success without appearing on our dashboard? Customers, junior staff, future maintainers, or the public often absorb what a metric excludes.
4. Disconfirmation: What observation would tell us that we are succeeding at the answer and failing at the problem? Decide before the results arrive.

This is not an argument against metrics, deadlines, or decisive answers. It is an argument for testing the target before celebrating precision.

A Sharper Question

Instead of asking:
“What happens when you answer the wrong question well?”

Ask:
“What result would look like success under our current question while proving we chose the wrong target?”

That question is harder to admire and harder to evade. It requires you to imagine the dashboard turning green at the exact moment reality turns red.

What to Do With This

At the start of a project, put the working question beside the human or business outcome it is meant to serve. For a hiring process, do not write only, “How do we fill the role faster?” Pair it with, “How do we identify someone who can perform the work and remain?”

In a review meeting, require one countermetric. If speed is celebrated, inspect rework. If sales rise, inspect consent, returns, and complaints. If an AI assistant gives a beautifully structured recommendation, ask it to state the assumption that would make its recommendation irrelevant or harmful before acting on it.

Finally, give someone permission to challenge the question, not merely the answer. The person who says, “Our analysis is excellent, but this is not the problem,” is not derailing the meeting. They may be preventing an elegant failure.

Bringing It Together

Answering the wrong question well does more than waste time. It makes error respectable, measurable, and repeatable. Before you improve the answer, inspect the target: what success would mean, who could be harmed by it, and what evidence would make you change the question. QuestionClass exists for that moment before momentum hardens into certainty. Practice one better question each day with QuestionClass’s Question-a-Day at questionclass.com.

Bookmarked for You

These books help you notice when a clean answer is serving a distorted goal rather than reality.

Thinking in Systems: A Primer by Donella H. Meadows - Meadows helps readers see how incentives, feedback loops, and narrow measurements can produce outcomes nobody intended.

The Scout Mindset by Julia Galef - Galef offers a practical case for seeing what is true rather than defending the result we want to be true.

Questions Are the Answer by Hal Gregersen - Gregersen focuses directly on how better questions disrupt familiar assumptions and open better routes to action.

QuestionStrings to Practice

A QuestionString is a short sequence that moves you from an attractive answer back to the reality it is supposed to improve.

False Target String
For when a project is producing impressive results but something feels off:

“What number or outcome are we currently celebrating?” →
“What larger purpose was that result supposed to serve?” →
“How could this result improve while the purpose gets worse?” →
“Who would notice the failure first?” →
“What should we measure or ask before continuing?”

Use this before approving a dashboard, strategy, AI recommendation, performance target, or process improvement. It is especially useful when everyone agrees that a result looks good but nobody has tested whether it means anything good.