MOAT Evaluator
In betaResearch
Codie Sanchez's MOAT framework, run by a panel of AI models instead of one.
The MOAT framework — Margin, Operations, Advantage, Total addressable market — is Codie Sanchez's test for whether a business is actually worth owning. MOAT Evaluator runs your idea through it and returns a score out of 40 with a FUND IT / FIX IT / FLEE IT verdict. It asks several AI models at once, because the useful signal is not the average score, it is where the models disagree.
What it does
The real framework, not a loose interpretation
Margin, Operations, Advantage and Total addressable market — the four tests Codie Sanchez uses to judge whether a business is worth owning, scored one to ten each, so you get a number you can compare between ideas.
A panel, not one opinion
The same evaluation is run by several models from different families at once, so one model's blind spot does not become your verdict.
Disagreement is the finding
Where the panel splits — say Margin scored 8, 5 and 3 — that dimension is flagged as contested rather than quietly averaged into a comfortable number.
Brutally honest by design
The panel is instructed to challenge your market-size claim, your acquisition story, your pricing and your scalability, and not to encourage you.
Risks and fixes, not just a score
Every evaluation returns the top risks and the changes that would most improve the score, so a low verdict is still actionable.
How a session goes
- 1
Describe the business idea — a paragraph is plenty.
- 2
The same MOAT prompt, built from Codie Sanchez's four dimensions, goes to several models at once.
- 3
Each returns scores out of 10 for Margin, Operations, Advantage and TAM, with reasoning.
- 4
You get the averaged score and verdict, plus the dimensions where the panel disagreed.
- 5
The full evaluation saves to your Library so you can compare ideas later.
What people use it for
- Pressure-testing an idea before you spend money or quit a job
- Deciding which of several ideas actually deserves your time
- Finding the weak dimension before an investor finds it for you
- Sanity-checking a plan a single AI already told you was brilliant
Which key it uses
MOAT Evaluator calls OpenAI, Anthropic (Claude), Google AI (Gemini) or xAI (Grok). On a paid plan, add a key for any of them under Account → API keys and that provider bills you directly — no credits spent. Without a key, runs spend from your included plan credits instead, and you are told when that happens.
Questions
- What is the MOAT framework?
- It is Codie Sanchez's four-part test for whether a business is worth owning, and it is what this tool scores your idea against. M is Margin: does it make real money after every cost, not just revenue. O is Operations: can it run and grow without you doing every task yourself. A is Advantage: do you hold something a competitor cannot easily copy — expertise, an audience, a licence, a distribution route. T is Total addressable market: are there enough buyers to make it worth the years. Sanchez popularised it for evaluating small businesses to buy or build, where a good-sounding idea with thin margins or no edge is the expensive mistake.
- Is this made by Codie Sanchez, or endorsed by her?
- No. MOAT Evaluator is built by Bizwax and has no affiliation with Codie Sanchez. We use her framework because it is a good one and she published it; the scoring, the AI panel and the verdict are ours, and any judgement the tool makes is ours to answer for, not hers.
- How close is the scoring to how Sanchez uses it?
- The four dimensions are hers and the questions under each one follow her intent — Margin looks at gross and net margin, acquisition cost and whether profit scales; Advantage looks for genuine founder-market fit rather than a slogan. The 1–10 bands and the FUND IT / FIX IT / FLEE IT verdict are ours. Treat the number as a way to compare ideas and find the weak dimension, not as a figure she would sign off.
- Why several models instead of the best one?
- Because scoring a business is subjective, and a single model gives you one confident number with no way to judge how much to trust it. If three models independently score Margin at 8, that is worth something. If they score it 8, 5 and 3, the honest answer is that it depends on assumptions you have not settled yet — and only a panel shows you that.
- Which keys does the panel use?
- Whichever provider keys you have saved. The more families you add — OpenAI, Anthropic, Google and xAI — the more meaningful the disagreement, because models from one family tend to share their blind spots. If you haven't saved a key for one of them, Bizwax runs that model for you instead, so you always get a full panel.
- What does it cost to run?
- On a paid plan, if you add your own provider API key the provider bills you directly and Bizwax charges no credits. Without a key, runs spend from the credits included with your plan — and the tool tells you when that happens. Organizations can supply one shared key for the whole team.
- Where does my work go?
- Everything you make is saved to your Library automatically, so it survives closing the tab. You choose what, if anything, to share to Creations.