Freeze what a good decision must accomplish.
Plinth distills your intent into numbered acceptance criteria, constraints, assumptions, and falsifiers. That contract becomes the reference point downstream roles cannot quietly rewrite.
START A DECISION BRIEF
Describe the decision.
Include the alternatives, organization, stakes, timing, constraints, and anything important the public internet will not know.
TRY A SAMPLE
ABOUT PLINTH
Research for a decision that is still open.
Describe the choice in your own words. Plinth searches for relevant companies, products, market moves, and constraints. You decide which comparisons belong in the brief. It then builds several credible ways forward and shows the case for and against each one.
Edit the research set.
The Researcher finds outside perspectives, but only the comparisons you choose can shape the brief. Context remains context; cited evidence remains evidence.
Challenge each move against the contract.
Each posture exposes tradeoffs, evidence, assumption load, criterion-level tensions, and one bounded next test. The reasoning trail ships with the brief.
A SMALL EXAMPLE
A software company is deciding how to enter a regulated market.
The brief might include this posture alongside two materially different alternatives.
Rent the regulated layer while demand is still uncertain.
- Why do it
- Reach the market sooner and learn which customer segments will pay before building fixed regulatory capacity.
- What it costs
- Lower margins, dependence on a partner, and less control over the customer experience.
- What must be true
- The partnership must preserve enough product differentiation to learn something transferable.
- Find out next
- Ask three qualified prospects to evaluate a partner-backed offer and commit to a paid pilot.
Good uses
Market entry, build-or-buy choices, product investments, partnerships, acquisitions, pricing changes, and operating-model decisions.
What to include
Deadlines, budget, internal capabilities, customer evidence, existing beliefs, political constraints, and anything the public record will miss.
What you will not get
A ranked answer, an ROI prediction, or an authoritative recommendation. The people responsible for the outcome still make the call.
WHY A BRIEF
The hard part is rarely a missing score. It is getting the actual disagreement onto the page without flattening it. A brief is small enough to read before a meeting and concrete enough to argue with during one.
WHO IT IS FOR
People preparing the room, not replacing it.
Product and strategy leaders
When several paths look plausible and the roadmap needs a defensible investment thesis.
Founders and operators
When time and capital are limited, but the available choices carry different kinds of risk.
Analysts and advisors
When the client needs a sourced starting point that can survive scrutiny and be revised in the room.
Leadership teams
When disagreement is present but buried inside different assumptions, vocabularies, or incentives.
DESIGN PRINCIPLES
The user keeps the wheel.
- 01
Reference intent before research
Numbered criteria are frozen before outside evidence can bend the question.
- 02
Context is not proof
What you supply is carried forward as context, not silently converted into external validation.
- 03
Separation of powers
Distiller, Researcher, Analyst, and Decision maker have distinct authority and cannot bless their own rewrites.
- 04
Unknown means unknown
Missing evidence becomes an assumption, falsifier, or next test—not a confidence score.
- 05
Reasoning ships
The audit trail is part of the artifact, not residue hidden in a system log.
PRACTICAL QUESTIONS
Before you begin.
How much context should I provide?
A paragraph is usually enough. Name the alternatives, why the choice matters, timing, constraints, and any internal facts that would materially change the analysis.
How long does the research take?
Usually between thirty and ninety seconds for competitor discovery, followed by another research pass when you build the brief. Source availability and decision complexity affect timing.
Does Plinth recommend an option?
No. It presents several defensible postures, their evidence, sacrifices, assumptions, counterevidence, and next tests. It does not rank them.
Can I change the competitors it finds?
Yes. Competitor selection is an explicit step. The final analysis uses the set you choose, not every result returned by the first search.
HAVE A DECISION IN MIND?
Start a decision brief ↑Put it on the table.
Plinth shows its work.
Sources stay attached. Assumptions remain visible. Unanswered questions do not quietly become facts.