What is a riskiest assumption test?

Test first the assumption that would kill the product if it were false, not the one that is easiest to test. The riskiest assumption is the one with the highest uncertainty and the highest consequence. Ordering experiments this way means you fail fast on the things that matter, instead of accumulating small confirmations of things that were never in doubt.

How do you prioritize what to test first in a startup?

Test first the assumption that would kill the product if it were false, not the one that is easiest to test. The riskiest assumption is the one with the highest uncertainty and the highest consequence. Ordering experiments this way means you fail fast on the things that matter, instead of accumulating small confirmations of things that were never in doubt.

Why test the riskiest assumption before the easiest one?

Test first the assumption that would kill the product if it were false, not the one that is easiest to test. The riskiest assumption is the one with the highest uncertainty and the highest consequence. Ordering experiments this way means you fail fast on the things that matter, instead of accumulating small confirmations of things that were never in doubt.

The riskiest assumption · Product Experimentation

§The question: of everything we believe, which belief, if wrong, ends the , and how sure are we of it?

§A team with a live product has a hundred things it could test. Most of them are easy, and easy is the trap. The order is not set by what is convenient to measure. It is set by what would be fatal to have gotten wrong.

§ 2.1Two axes#

§Every assumption sits on two axes. Consequence: if this is false, how much does it cost, from a wasted afternoon to a dead company. Uncertainty: how sure are we, from proven to pure guess.

§The is high on both: we are not sure, and if we are wrong it is fatal. That is the one to test first, because a confirmation there de-risks the whole product, and a refutation there saves everything you would have spent building on top of it.

§David Bland and Alexander Osterwalder built a whole method on this single move, the riskiest-assumption test, and the reason it needs a whole method is that every instinct pulls the other way. The easy assumptions are pleasant to test and produce a steady stream of small green results that feel like progress and de-risk nothing.

What teams test first	What they should test first
Button color, copy tweaks, layout: easy, low consequence, low uncertainty.	Whether recognized guests actually visit more: hard, fatal if false, and genuinely unknown.
Whether users notice the new feature.	Whether users who notice it come back next month.

§ 2.2The fatal cluster#

§Marty Cagan's four risks name where fatal assumptions usually hide: value (will they use it), usability (can they), (can we build it), viability (does it make money). After launch, feasibility and usability are largely answered, the product exists and people can use it, so the riskiest assumptions cluster in value and viability: whether the thing delivers the value the evidence promised, and whether the economics work when real money is involved.

§For most first versions, the single riskiest assumption is retention: whether people come back. It is high consequence, because nothing works without it, and high uncertainty, because pre-launch evidence cannot prove it. That is why Part II puts retention near the front.

§ 2.3Failing fast where it counts#

§Rita McGrath's discovery-driven planning made the case thirty years ago and it holds after launch: test the assumption that kills the plan first, so that if the plan is dead you find out while it is cheap. "Fail fast" has become a slogan that teams apply to trivia, failing fast on button colors while the fatal assumption goes untested for a year. The phrase only means anything applied to the riskiest assumption. Failing fast on things that do not matter is just being busy.

§ 2.4What you leave with#

§The assumptions from the running list, plotted on consequence against uncertainty. The one in the top-right corner, named, as the next experiment. And an honest note when the top-right corner is being avoided because its likely answer is unwelcome, which is the most important thing to catch and the hardest.

2. The riskiest assumption