Exploratory Prototype¶

Artifact or experiment — instantiates Curiosity Gap Design

Implements the first probe by creating a lightweight artifact, scenario, mockup, or pilot that tests what is unknown.

Version: v1 · 2026-08-24 · History
Mechanism #: 3430
Type: Artifact or Experiment
Form family: Experiment, Test & Rehearsal
Solution family: Optimization & Search
Problem family: Agency, Participation & Relational Trust Failure
Problem subfamily: Inquiry & Epistemic-Openness Failure
Origin domain: Engineering & Design
Also from: Human-Computer Interaction, Innovation & Entrepreneurship
Instantiates: Curiosity Gap Design

Some unknowns can't be talked into the open — you have to build a cheap thing and watch what it does. Exploratory Prototype is a deliberately lightweight artifact — a mockup, a scenario, a foamcore model, a weekend pilot — built for one purpose: to answer a specific learning question about an unknown. Its defining move is that the artifact exists to be interrogated, not to be kept: it is constructed only as far as the question demands, and its feedback is read as evidence about that question, never as validation of the finished thing. It differs from a Discovery Task, where participants generate evidence by acting in the world with no artifact to build, and from a Guided Exploration Path, which routes people through existing material. Here the team manufactures a disposable test-object precisely because the answer isn't yet available anywhere to be found.

Example¶

A city transportation department suspects a protected bike lane would work on a busy commercial street, but pouring concrete is expensive and nearly irreversible, and nobody can say for certain whether riders will actually come or car traffic will seize up. Instead of committing, they run an exploratory prototype: over one weekend they lay down a "pop-up" protected lane with planters, traffic cones, and paint — tactical urbanism, cheap and removable. The learning question is written on the work order before a single planter is placed: does a protected lane on this street raise weekday cyclist counts without pushing car travel time past two minutes of added delay? For two weeks they count bikes and clock cars. The feedback lands: cyclist volume roughly triples, car delay adds about forty seconds. Crucially, the team treats this as evidence toward the question, not proof the permanent design is right — the pop-up can say nothing about snow clearing, delivery loading, or the parking-loss politics a permanent lane will trigger. So the prototype earns the next probe, a longer semi-permanent pilot, rather than being mistaken for a finished verdict.

How it works¶

Name the learning question first. Write down the single unknown the artifact must answer before building, so that whatever the prototype produces is interpreted against that question rather than reverse-justified afterward.
Build the cheapest thing that can answer it. Fidelity rises only to what the question needs — cones and paint, a fake front-end, a paper mockup. Over-building squanders the whole advantage of a probe, which is that it's cheap to be wrong.
Expose it to real conditions. The artifact has to meet actual users, actual traffic, actual reality; feedback gathered in a sandbox that flatters the design is theater.
Read feedback as evidence, not verdict. Capture what the test says about the question and, explicitly, what it cannot say — the guard against a scrappy test hardening into a shipped commitment.

Tuning parameters¶

Fidelity — how realistic and polished the artifact is. Higher fidelity yields more trustworthy feedback but costs more and quietly biases everyone toward committing to what they've already built.
Question scope — one unknown per prototype versus several bundled together. Narrow probes give a clean read; bundling muddies which feature actually drove the result.
Exposure realism — private sandbox versus live in the world. More realistic exposure surfaces true friction but risks real consequences if the prototype fails badly.
Reversibility — how disposable the artifact is. Fully throwaway protects against sunk-cost attachment; a semi-permanent pilot blurs into the real thing and gets hard to tear down.
Observation horizon — how long you watch before reading the result. Longer captures durable effects (novelty wearing off) but delays the next loop.

When it helps, and when it misleads¶

Its strength is converting an argument about "will it work" into cheap evidence, and de-risking a big commitment before the money is spent — the build–measure–learn logic in which every artifact exists to test one assumption and earn the next.^[n1] A good prototype is the lowest-regret way to buy down a genuine unknown.

Its failure mode is the prototype that quietly becomes the product: feedback treated as validation, and a weekend hack hardened into permanent infrastructure without ever being re-tested at real fidelity. A close cousin is the rigged demo — an artifact built to confirm a decision already made rather than to learn, which produces applause and no information. The classic misuse is "the pilot went great, so ship it," when the pilot answered a much narrower question than the shipping decision actually requires. The guarding discipline is to keep the learning question fastened to the artifact and to state plainly what the prototype's feedback cannot certify — so the next probe, not premature rollout, is what the result earns.

How it implements the components¶

first_probe — the artifact is the initial, low-cost investigating move into the unknown; building and exposing it is the first step of inquiry.
relevance_link — every prototype is pegged to an explicit learning question tied to a real decision, so its feedback matters rather than being idle exploration.
answer_capture — the prototype's output is recordable feedback about the question, captured together with a note on what it does and does not show.

It does not stage a rising sequence of questions — that is Guided Exploration Path and Curiosity-Driven Onboarding Path — nor does it maintain the cumulative record across probes, which is Inquiry Log's job.

Instantiates: Curiosity Gap Design — the prototype is the built first probe that turns an unknown into evidence.
Consumes: Provocative Question Prompt or Research Question Workshop — supplies the explicit learning question the artifact is built to answer.
Sibling mechanisms: Provocative Question Prompt · Mystery Frame · Guided Exploration Path · Research Question Workshop · Discovery Task · Curiosity-Driven Onboarding Path · Inquiry Log

Editorial Notes¶

Form Classification¶

Form family: Experiment, Test & Rehearsal

Rationale: Exploratory Prototype operates as a bounded trial, probe, simulation, or rehearsal that generates evidence from performance because it implements the first probe by creating a lightweight artifact, scenario, mockup, or pilot that tests what is unknown.

Independent corroboration: The frozen evidence defines Exploratory Prototype as 'Implements the first probe by creating a lightweight artifact, scenario, mockup, or pilot that tests what is unknown', so its operative form is Experiment, Test & Rehearsal.

Review outcome: Independent reviewer agreement; high confidence.

Origin Attribution¶

Primary origin: Engineering & Design

Origin pattern: Convergent development

Present-day reach: Multi-domain

Rationale: Building low-cost artifacts to test unknowns is a longstanding engineering and design prototyping method.

Related originating lineages:

Human-Computer Interaction — Interaction design independently developed paper and interactive prototypes for user learning.
Innovation & Entrepreneurship — Lean experimentation materially shaped prototypes as uncertainty-reduction probes.