Research & Discovery
The knowledge quadrant: sort evidence from assumption
Before committing to build, sort every claim into known, assumed, or unknown, and buy research only where assumption meets risk.
Most project risk hides in unexamined assumptions. The knowledge quadrant, run at the start of discovery, sorts every claim into what is known on evidence, what is assumed, and what must be researched, so teams stop building the wrong thing confidently.
At the start of a project, a team's 'knowledge' is a mix of three things wearing the same clothes: facts backed by evidence, assumptions that feel like facts, and open questions nobody has voiced. The knowledge quadrant separates them. Every claim about the customer, the problem, and the market gets sorted: known with evidence, assumed, or still to be discovered. The map shows exactly where the team is shooting in the dark and where observational research is required before commitments get made. The real output is not the map; it is a research plan and a shorter, safer list of commitments.
Why it matters to the business
Confident assumptions are the most expensive kind of wrong. Bain found 80% of companies believed they delivered a superior experience while only 8% of their customers agreed; inside views inflate reliably. Even direct customer claims need checking: a sampling study by the analytics firm 84.51 found 75% of respondents misstated their own purchase behavior when compared with loyalty-card data. Forrester's journey-mapping guidance draws the same line: draft your hypothesis from internal insight if you must, but never publish the assumption map as truth. The quadrant is that discipline applied to a whole project.
How to use it
- Run the quadrant at kickoff, before scoping. Sorting claims after commitments are made just decorates the decision.
- Demand a source for every 'known': who observed it, when, with which customers.
- Convert the riskiest assumptions into research questions with a method, an owner, and a deadline.
- Validate what customers say against what they do. Behavioral data beats stated intent.
- Define out-of-scope items early, before the feature factory fills the gaps with guesses.
- Revisit the quadrant at every phase gate; yesterday's unknown may now be evidence.
Where teams get it wrong
Workshops manufacture false knowledge. A guess gets written on a sticky note, voted on, repeated in three meetings, and graduates into a fact, though no customer ever saw it. The quadrant fails the same way when it is run once, filed, and never updated as evidence arrives.
Ask your team
- For our biggest current bet, which claims rest on evidence and which on assumption?
- What finding would kill this project, and have we looked for it?
- When did a real customer last contradict something we believed?
A guess written on a sticky note is still a guess.
Apply this
Reading about the knowledge quadrant: sort evidence from assumption is one thing. Seeing where it applies in your journey is the useful part.