Prioritising Assumptions

A thorough assumption-mapping exercise might reveal 20, 30, or even more assumptions for a single feature. It might feel safer to test everything, but we don’t have time for that, so this means that we need to be strategic about which assumptions we validate first. But how do you compare a desirability assumption against a usability or viability one?

Step 1: Rank Importance

The first step in prioritisation is determining how critical each assumption is to the success of your feature or product.

Ask yourself: If this assumption proves false, how much will it impact the overall solution?

• High Importance: Assumptions that underpin the core functionality of your product. If these fail, the entire feature is at risk. For example, if your product assumes that users will willingly share their data to access personalised recommendations, this is a critical behavioural assumption.

• Low Importance: Assumptions that, if incorrect, may cause minor inconveniences but don’t fundamentally undermine the product.

This ranking helps separate "must-validate" assumptions from "nice-to-validate" ones.

Step 2: Assess the Level of Evidence

The second dimension is evidence. For each assumption, consider how much confidence you have based on existing data, research, or prior experience.

• Low Evidence: Assumptions based on gut feelings, anecdotal feedback, or untested hypotheses. These are high-risk because there’s little to support their validity.

• High Evidence: Assumptions grounded in extensive research, historical data, or established patterns. For example, if past user tests consistently show that customers understand and prefer a specific design pattern, you can be more confident in its usability.

The less evidence you have, the more important it is to validate the assumption. There isn’t a definitive scale for what constitutes low versus high evidence so you need to use the judgement of the team when defining your position.

With the importance and evidence levels established, plot your assumptions on a 2x2 matrix. This visual approach helps you to quickly identify which ones are critical, leap-of-faith assumptions (high importance, low evidence) and which can be deprioritised (everything else).

Map and test multiple solutions simultaneously

The outcome of a good ideation session is, at a minimum, 3 prioritised solutions because the chances of any given solution achieving our goals is low. Given the high level of uncertainty, we need to test multiple solutions simultaneously to increase our chances of success.

This means that you need to map the assumptions for all three solutions at the same time. By visualising everything on the same prioritisation chart you can quickly see which assumptions are critical for each solution and which assumptions are shared across all solutions.

We should always test two or three assumptions in parallel, using very quick and cheap experiments, because we need to move quickly through to the next phase. In selecting the first assumptions, we should focus on the leap-of-faith quadrant (high importance, low evidence). But if you have quite a few assumptions to validate in that quadrant you will need to prioritise them.

Start with assumptions that relate to more than one solution, even if there is a higher importance or more uncertain assumption. If we can invalidate all of our solutions quickly we can go back to our ideation and start again with our newfound knowledge. With a selected of 3-5 assumptions you are ready to move to the next stage: evaluation.

Conclusion

In product development, every feature is built upon a set of assumptions. Rather than trying to test a feature as a whole, we can save a lot of time and effort by validating the core assumptions without which the solution would fail. A good assumption mapping exercise will yield dozens of assumptions, but we can’t verify them all so we need a way of prioritising the most critical assumptions.

By ranking assumptions based on importance and evidence and mapping them across multiple solutions, we can focus our efforts on the high-impact, high-uncertainty areas. Using this prioritised list we can efficiently validate multiple assumptions and determine whether our solutions are viable or not.