InnoveraInnoveraThe Perfect Toaster

Project Frame

The shared brief for The Perfect Toaster. What this project is, why it exists, and what we're trying to accomplish. Read this first.

The case, in Q & A

What kind of project is this?

This is a consumer-hardware project. We are designing and bringing to market a physical product, a better everyday toaster, and using this workspace to run the whole effort in the open: research, design debates, planning, and the sources that back every decision.

It is a collaborative case, not a solo build. People join, weigh in on the questions that matter, and the project moves as the team converges. Leo watches the whole thing and surfaces the next useful step.

What problem are we trying to solve?

Toasters have barely changed in decades. They burn one slice while under-toasting the next, the settings are guesswork, and the good-looking ones are expensive while the affordable ones look like appliances from a rental unit.

We want to settle what a genuinely better toaster is, even toasting, controls people trust, and a form that earns a spot on the counter, and prove there is real demand for it before we commit to tooling.

What do we want to accomplish here?

Reach a shared, evidence-backed answer to three things: who this is for, what it must do, and whether the numbers work. Each of those lives in the workspace. The market read, the design direction, and the execution plan, and the Frame is the brief that ties them together.

Concretely, by the end of this phase we want a validated design direction, a demand signal we believe, and a plan with a credible path to a first production run.

How do we know when we're done?

We are done framing when the team can state the pitch in one paragraph without argument, the top open questions have owners, and the summary reads as a decision rather than a debate.

From there the project graduates from framing into build. The plan takes over and the Frame becomes the reference we check ourselves against.

How should the team use this page?

Treat this as the source of truth for intent. If a decision elsewhere in the project seems to drift from what's written here, that's a signal to either change the decision or update the Frame, deliberately, together.

New collaborators should read this first. It's the fastest way to understand what we're doing and where they can help.

The subject, and its boundaries

What is modelled is a position in premium counter-top appliances, entered through a repairable toaster. Not the toaster alone: the toaster is the first product, and the model has to survive the question of whether it should be the only one.

Inside the boundary: the design-led premium tier in the US, direct and specialty retail, the consumable and repair line. Outside it: commercial and food-service, non-US markets before 2028, and any product requiring firmware. Those are boundaries rather than decisions not yet taken, and moving one is a frame amendment.

Questions this model must answer, and over what horizon
Question classInstanceHorizon
ProceedDo we commit tooling capital at 5K unitsnow, first use
SequenceCampaign first or certification first6 months
AllocateSecond SKU or deeper into this one18 months
RepositionDoes the gifting buyer or the self-purchaser lead12 months
ExitWhat would make open-sourcing the design the right answerany time
Fidelity required, by dimension
DimensionFidelityWhy that level
Unit economicsdecision-gradeThe tooling commitment turns on it
Demanddecision-gradeEverything downstream inherits the volume assumption
CompetitiveworkingEnough to know the white space is real, not to war-game it
Operations and legaldecision-gradeCertification timing gates revenue
TeamsketchSix people. The gaps are visible without modelling
Financial modelworkingThe three-year view informs; the trough decides
Objectives, and constraints that genuinely bind
Objective: a product that is bought once and keptMeasured by element attach and return rate, not by units. This is the objective the model optimises.objective
Constraint: $240K, no outside capital before the campaignBinds. It is what makes September 2026 the decision point rather than a milestone.binding
Constraint: certification before revenueBinds absolutely. Not relaxable at any price, which is why it sits on the critical path.binding
Constraint: assembly stays domestic for run oneDoes not bind. Stated as a preference and testable against the second-source quote.slack
How fixed is the solution

Flexible. The repairable premium toaster is the current answer, not a given: the frame commits to the opportunity, which is the gap between what the category sells and what a slice of buyers say they want, and the two-slot mechanical toaster is a hypothesis against it.

This matters for what Opportunity Validation is allowed to conclude. Under a fixed solution its job would be to find the best route to market for this toaster. Under flexible, it is allowed to come back and say the toaster is the wrong entry point, and that verdict would be a finding rather than a failure.

Settled before analysis, and not open to relitigation
No firmware, no connectivityA positioning commitment. Reversing it forfeits the argument the product is making.settled
Repair documentation published openlySettled June. The moat is brand and supply chain, not secrecy.settled
Two-slot formatLocked Jun 25. Revisit only if demand data flips.settled, conditional
Community co-design in publicPredates the frame. It is how the project works, not a channel choice.settled
What triggered this, and what intake surfaced

The triggering question was narrow: should we commit to tooling at 5K units. That is recorded as first use rather than as the definition, because a model built only to answer it would be discarded the moment it was answered.

Three questions intake raised that had not been asked. Whether the buyer is the user, which turns out to be unevidenced and load-bearing. Whether the repair line is a business or a marketing position, which changes what the product has to be. And what would make stopping the right answer, which had no agreed answer at all until the kill case was written.