The Cheapest Question That Could Prove You Wrong — Altvina Altvina Insights

Published July 15, 2026 · Altvina Insights · 5 min read

The Cheapest Question That Could Prove You Wrong

Move three, the one we wouldn't skip. Before you spend a dollar fixing your problem, find the cheapest question that could show your guess is wrong.

Before you approve any changes, pause. Take your best guess about what's wrong, the one you feel pretty sure about, and ask: what's the simplest thing you could check that would prove the guess wrong? Not what would confirm your theory, but what would disprove it. Then go test that, and do it before you reorganize anything or hire anyone. Before you buy the tool, even. Of all five moves, this is the one we'd defend, because it comes right before the spending starts. That's where costly mistakes happen.

This is the third move in the diagnosis we've been running live all week. (The firm, a reminder, is a composite: patterns from firms this size, not a client we could name, and no real quotes.) Two moves in, we have a suspect. We noticed work stops at one review with no fixed slot in the schedule. We found the loud complaint, the sales process, gets worse when sales improve, meaning it's a symptom. The suspect is the stalled review. But a suspect is just a guess, and today is about testing that guess before you invest in it.

Why this is where diagnosis goes wrong

In our experience, this is where diagnosis usually goes wrong. You can observe the work closely and separate the complaint from the constraint honestly, yet still end up focusing on the wrong thing. Our minds are designed to collect evidence for what we already suspect. By the time a founder has a theory, every meeting seems to confirm it. The theory feels more certain by the day, and that certainty is often just the same guess, repeated.

The disconfirming question is the fulcrum of the whole method. It interrupts that pattern, and it’s intentionally cheap. You’re not trying to prove your theory; you’re giving reality a chance to embarrass you before you spend money. If it doesn’t, your guess holds up, and you can spend with confidence. If it does, you’ve learned for the price of a question instead of a costly reorganization.

The two-week test

Our suspect is the one review every proposal must pass through. The expensive fix the founder is considering is hiring someone to handle the overflow and keep things moving. It’s reasonable, but it could mean a year’s salary spent on a theory.

So before any of that, we ask the cheapest disconfirming question. If the bottleneck really is that single review, removing it for two weeks should let deliverables ship on time. That’s the test, and it costs almost nothing if you cap the risk: limit the batch to the smaller, routine engagements, and have the founder spot-check pricing within a day of send instead of gating each one up front. The queue disappears; the safety net stays. Let the senior associate, who’s already idle, scope and price that batch of proposals without routing them through the founder. Then watch whether the proposals actually go out on time over those two weeks.

There are only two outcomes, and both are useful. If deliverables start shipping on time, the guess is confirmed cheaply, and any fix becomes a justified decision instead of a hopeful one. If they’re still late, even without the review, the guess was wrong. The real issue is elsewhere, and the founder just saved a year’s salary on a fix that wasn’t necessary. Either way, you learn the truth for the cost of two weeks.

Notice the shape of the question. It’s small, it’s fast, it’s reversible, and it’s aimed squarely at the thing that would hurt your theory most. A good disconfirming test isn’t the one that’s easiest to run. It’s the one your guess is most afraid of.

How to pick the cheapest test for your own guess

Take the suspect you identified yesterday and match it to the cheapest test that could prove it wrong.

If your suspect is that everything waits on your sign-off: pick one low-stakes category and skip your review for two weeks; see whether those items shipped and whether anything actually went wrong.

If your suspect is that you’re underpriced: pull the real margin on your last ten jobs, one number most firms have never actually computed, before you touch the rate card.

If your suspect is one person: move the task to someone else for two weeks and see whether the stall moves with the person or stays with the step.

Each is reversible, costs days not quarters, and is aimed at the version of reality that would embarrass the guess.

Run yours, and you’ve done the genuinely hard part of diagnosis that many firms skip on their way to spending. They jump straight from suspicion to budget. The cheap disconfirming question is the stop in between, and few of us take it.

This is move three of five. The full method, all five moves laid out, is on a single card we’re sharing today. Save it if it helps; that’s its purpose. The last two moves, pinpointing the one constraint that shifts the others and writing the honest recommendation, come Thursday and Friday. For now, focus on the one we’d protect above the rest. Before you pay to solve your problem, ask the cheapest question that could prove your guess wrong, and ask it while being wrong is still cheap.

This week's diagnosis, one move a day: JSX0 · JSX1 · the cheapest question that could prove you wrong · JSX2 · JSX3

Continue the series

This is part 3 of a 5-part series on Watch the Diagnosis. The full arc:

Content and Accuracy Disclaimer

This article was drafted with AI assistance and reviewed by the Altvina team. We rigorously fact-check all content to ensure reliability.

Should you notice any inaccuracies or outdated information, please contact us so we can correct it. Your feedback helps us maintain high standards of accuracy and transparency.