--- title: "The 7 Powers: How to Score a Company's Moat (2026)" description: "A plain English guide to the 7 Powers framework: what each power means, how to score a company's moat, and what to ask an AI before you trust the answer." url: https://moneda.com/blog/what-are-the-7-powers lang: en --- # The 7 Powers: How to Score a Company's Moat (2026) > A plain English guide to the 7 Powers framework: what each power means, how to score a company's moat, and what to ask an AI before you trust the answer. Published: 2026-10-23 Updated: 2026-10-23 Author: Federico Kunze Küllmer Source: https://moneda.com/blog/what-are-the-7-powers --- *Federico Kunze Küllmer is the founder of Moneda, one of the tools compared here.* > **Short answer**: The 7 Powers is a framework from strategist Hamilton Helmer for judging whether a company can keep earning more than its rivals for years. A company has "Power" when it gets a real benefit (higher prices or lower costs) and competitors face a barrier that stops them copying it. The seven sources are Scale Economies, Network Economies, Counter-Positioning, Switching Costs, Branding, Cornered Resource and Process Power. To score a moat, check each power against the company's filings, write down the evidence, and be strict: most companies have one or two powers at most, and many have none. A moat is the thing that protects a company's profits from competition, the way a moat protects a castle. Everyone says the word. Fewer people can say which moat a company has, or how they would know if it was drying up. The 7 Powers gives you a checklist for that. This guide walks through each power, shows you how to score it, and lists the questions worth putting to an AI tool, plus what you still need to check yourself. If you want to see the framework used on real companies, the [Acquired](https://www.acquired.fm/episodes) podcast closes each company episode with a 7 Powers section. The hosts go through the powers one by one and say which ones the company has. ## What counts as a "Power" Helmer's test has two parts, and a company needs both: 1. **A benefit.** The business gets something that lifts its cash flow: it can charge more, spend less, or invest with less risk. 2. **A barrier.** Competitors cannot take that benefit away, even if they want to and have money to try. A benefit with no barrier gets competed away. A great new product is a benefit. If a rival can ship the same thing in eighteen months, there is no barrier, so there is no Power. Keep this in mind the whole way through. Most bad moat analysis skips the barrier. ## The seven powers, one by one | Power | The benefit | The barrier | A quick test | |---|---|---|---| | Scale Economies | Unit costs fall as volume grows | A smaller rival can't reach that volume without big losses | Do costs per unit drop as revenue rises? | | Network Economies | The product gets more useful as more people use it | Users won't move to a smaller network | Would a new user pick the biggest network by default? | | Counter-Positioning | A new business model beats the incumbent's | The incumbent won't copy it, because copying would hurt its existing business | Would copying cost the leader its current profits? | | Switching Costs | Customers stay and pay more over time | Leaving costs them time, money or risk | What does a customer lose by switching? | | Branding | Customers pay more for the same thing | Trust builds over years and can't be bought quickly | Would people pay more if the logo were removed? No? Then branding is doing the work | | Cornered Resource | Exclusive access to something valuable | Others can't get it on fair terms | Patent, licence, unique talent, key location? | | Process Power | Better costs or quality from how work is done | The know-how is spread across the organisation and takes years to copy | Have rivals tried to copy the process and failed? | Scale Economies, Network Economies, Switching Costs and Branding are the ones you'll meet most often. Counter-Positioning tends to explain how a challenger grows. Process Power is rare, and people claim it far more often than it exists. ## How to score a company's moat, step by step ### Step 1: Get the primary documents Start with the annual report. For US companies that's the [**10-K**](/blog/how-to-read-a-10-k-with-ai), the yearly filing every listed company sends to the [US Securities and Exchange Commission](https://www.sec.gov). Read the business description, the risk factors and the segment data. Add the last few quarterly reports and an earnings call or two. You're after facts: market share, customer counts, retention rates, unit costs, patents and their expiry dates, major contracts. ### Step 2: Check whether there's anything to protect A moat only matters if the company earns good returns. Look at operating margins over five or more years. Look at **free cash flow**, the cash left after running the business and paying for new equipment. Look at return on invested capital. If margins are thin and falling, scoring powers is mostly academic. Something is already getting through. ### Step 3: Check each power for a benefit and a barrier Go through the seven powers one at a time. For each, ask the two questions from Helmer's test: is there a real benefit, and is there a barrier that stops rivals copying it? Write one line of evidence from the filings next to every answer. If you can't find evidence, treat the power as absent. ### Step 4: Look for the barrier failing For each power you found, ask what would break it. A patent expires. A new technology makes scale matter less. A regulator forces data portability, which can cut switching costs. A brand gets hit by a scandal. Write down the one or two threats that would hurt most, and what you'd expect to see first if one of them started happening. ### Step 5: Connect the moat to the price A strong moat doesn't make any price a good one. The **P/E ratio** (share price divided by earnings per share) tells you how much the market already expects. A [**DCF**](/blog/how-to-value-a-stock-with-ai), or discounted cash flow model, estimates what future cash is worth today. If the price assumes the moat lasts forever and your evidence says ten years, the moat is fine and the price is the problem. ### Step 6: Revisit it Moats change slowly, then quickly. Check it again once a year after the new 10-K, and sooner if you see one of the warning signs from Step 4. ## Example questions to ask an AI [AI tools](/blog/best-ai-stock-research-tools) save a lot of reading time here, as long as you ask for evidence and not opinions. Some questions that work well: - "Which of the 7 Powers does this company have? Quote the line from the latest 10-K that supports each one." - "Which of the 7 Powers does this company not have? Explain why." - "Has gross margin gone up as revenue grew over the last five years? Show the figures and where each one came from." - "What would a competitor need to spend to match this company's scale?" - "Give me the bull case and the bear case for this moat lasting ten more years." - "What changed in the risk factors section between the last two annual reports?" [Moneda AI](/financial-agent), the agent inside the Moneda app, can assess a company's moat for you. It works from company filings and live market data and shows the source behind each figure. It also gives the bull and bear case. It describes. It never tells you to buy or sell. The honest limit: Moneda covers US-listed shares and ETFs (funds that trade like a single share) only, so no European-listed stocks or funds. If you use [ChatGPT](https://chatgpt.com), [Claude](https://claude.ai) or another general chatbot, the same questions work. Just paste in the filing or point it at the source, and ask it to cite. | | Moneda AI | General chatbot (ChatGPT, Claude) | |---|---|---| | Where the data comes from | Company filings and live market data | Whatever you paste in or it can find | | Sources | Shown behind every figure | Only if you ask it to cite | | Invest after the research | Yes, in the same app, with your approval on every trade | No | | Coverage | US-listed shares and ETFs only | Any company you give it documents for | | Best for | Answers that show the source behind every figure, then investing in the same app | Ad hoc reading of a filing you already have | ## What to verify yourself Even when an answer cites its sources, check these by hand: - **The quote is real and current.** Open the filing and find the line. Make sure it comes from the latest report and not one from three years ago. - **The numbers add up.** Recalculate one or two margins yourself. If a single figure is wrong, treat the rest with care. - **The barrier, not just the benefit.** Did the answer explain why rivals can't copy the advantage? If it only described the advantage, push back. - **Who the competitors are.** Check that the AI compared the company with its real rivals, including private ones and new entrants. - **Recent events.** Look for lawsuits, regulation or product launches from the last few months that might not show up in the filings yet. ## Putting it to work in Moneda If you want research and investing in one place, Moneda AI is built on [Numa 1](/models/numa-1), Moneda's financial AI engine. It researches stocks, ETFs, gold, silver and oil (through ETFs) and crypto. You can describe an idea in plain language, such as "software companies with high switching costs and growing free cash flow", and it finds candidates, or you can filter with the stock screener. Its search covers SEC filings and earnings-call transcripts, so you can check what management says about pricing power or switching costs in its own words. From there it can build an illustrative portfolio from your goal, track it, tell you when something changes, [stress test it](/blog/best-ai-tools-to-stress-test-a-portfolio) against 2008, 2020 and 2022, and compare it with gold, an index or any fund. You can [buy US listed shares](/invest) and ETFs (including fractional shares) and crypto in the same app. Every trade is approved by you with your passkey, using your face or fingerprint. For accuracy, Moneda's own tests in October 2026 measured 97.9% groundedness, 92% claim verifiability, 95.6% on FLARE-CFA, and 80% on the public set of the [Vals AI](https://www.vals.ai) Finance Agent Benchmark v2. That 80% run was done and graded by Moneda. Vals AI's top published score is 65.4% on its private set, which is a different set, so the two numbers aren't a direct comparison. [Standard is free](/plans). Premium starts at €9.99 a month and Ultra at €19.99 a month, with more AI usage and 1, 5 or 10 commission free trades a month depending on plan. If you'd rather work in Claude or ChatGPT, [Moneda MCP](/developers) connects them to your Moneda account with the permissions you set, and payments are still approved in the app. The REST API and CLI come with Premium and Ultra. ## FAQ **Who created the 7 Powers framework?** [Hamilton Helmer](https://www.strategycapital.com), a strategy consultant and investor, set it out in his book [*7 Powers: The Foundations of Business Strategy*](https://7powers.com). **What's the difference between a moat and a Power?** They mean nearly the same thing. "Moat" is the everyday word. Helmer's "Power" is stricter: it needs both a benefit and a barrier that stops competitors taking the benefit away. **How many powers does a good company usually have?** Often one, sometimes two. A company with three or more strong powers is rare. If your analysis finds five, go back and check your evidence. **Can a small company have a Power?** Yes. Counter-Positioning and Cornered Resource often show up early, before a company has scale. Scale Economies and Network Economies usually come later. **Can AI score the 7 Powers reliably?** It can do the reading fast and point you to the evidence. The judgement is still yours. Ask for sources, check a few of them, and be wary of any score that comes without one. **Does a strong moat mean the stock is a good buy?** No. A great business can still be expensive. Compare what the price assumes with how long you think the moat will hold. *This guide is for education only and is not investment advice.*