Glossary

What is assumption testing?

Assumption testing is the practice of writing down what must be true for an idea to work.

You rank those beliefs by how risky and how unknown they are, then design the cheapest possible test for the riskiest one.

Assumptions fall into four buckets: desirability (do they want it), viability (does it make money), feasibility (can we build it), and usability (can they operate it). The ones worth testing sit in the top-right of an assumption map: high damage if wrong, and no evidence yet.

Why it matters

Teams test what is easy to test, which is usually feasibility, because engineers can answer it. The assumption that kills products is almost always desirability, and you can answer it in a week of conversations for a fraction of the cost of the build.

Example

Assumption: "consultants will pay for this out of their own pocket." The cheapest test is asking six consultants what they pay for out of pocket today, and for what. Two hours of calls, and it can end the project.

In UpSight

You write your assumptions into the project setup, and UpSight uses them to draft the decision questions and research questions the project tracks. As interviews come in, evidence is linked to those questions automatically, and each question shows what has been answered and what is still open. There is no separate assumptions board with a validated or invalidated status. The question list is where that record lives.

Related terms

FAQ

What is the riskiest assumption test?
The cheapest experiment that could disprove the belief most likely to kill the idea. Cheapest and most fatal, in that order.
How do you prioritize assumptions?
On two axes: how much damage if the assumption is wrong, and how much evidence you have now. Test high-damage, low-evidence first.
Can you test assumptions without building anything?
Usually, yes. Conversations, landing pages, concierge delivery, and manual prototypes answer most desirability assumptions.