Decision Compression
A model for guiding users from vague intent to decision without requiring them to specify what they want in advance.
Developed Mechanism Established
Human Behaviour
Most people do not know exactly what they want until they see it. Asking them to describe it first produces a best guess, not a real preference. Systems built around that guess stall the decision rather than drive it.
Mechanism
Instead of leading with a question to unearth more, the system leads with a proposal while using existing and contextual knowledge. It shows a set of options, consisting of confident recommendations, popular alternatives, and deliberate contrasts. The set acts more as a diagnostic than a shortlist. Each response, whether acceptance, rejection, or passing interest, narrows what comes next without asking the user to explain why. People know what they do not want when they see it, even when they cannot articulate their wish in advance. Across a few turns, the right answer emerges.
How to know it worked
The person stops browsing and performs an action. Any reaction to the first set of proposals, either diving deeper into the set or a rejection and pivot, gives an early sign that the mechanism is working. A committed action is confirmation it worked.
Where it breaks
The model fails when inventory is too thin to support more than one or two rounds as the system runs out of fresh options before the user finds a base. Where users already know what they want and find the back-and-forth slower than a direct search, the latter with needle-point queries might serve better. It also breaks when users game it, pushing back not because the options are wrong but to see what else surfaces.
What makes it distinct
Decision Compression is not a filtering model. The user does not arrive with declared constraints: they emerge from the conversation. The nearest adjacent model, Elimination Sequencing, starts from the opposite assumption: the user already knows their limits and needs help applying them. Elimination Sequencing induces preference, while Decision Compression extracts it.