Porter’s Five Forces Answers a Different Question

Founders reach for Five Forces to decide whether to enter a market. It was never built to answer that question.

A seed team enters with a product that is genuinely better, and early users agree. Then a platform changes its rules and acquisition costs double. Buyers discover that switching is cheap and start negotiating. A copy appears within weeks. Nothing went wrong with the product. What went wrong is that other people held the leverage, and leverage gets paid for out of margin.

Five Forces would have shown that coming. Not as a verdict on the market, which is what founders usually want from it, but as an account of who had a claim on the money before the team arrived.

A venture spends its early life searching for a business: who it sells to, what it sells, how it makes money. That search runs inside a structure it did not choose and cannot see from the inside. Reading the structure is a separate skill from running the search, and it answers a narrower question than most founders ask of it.

What the model is actually for

Porter was not writing for founders. He was answering an economist’s question: why do some industries stay more profitable than others, decade after decade, when competition is supposed to erode exactly that? His answer was that five structural conditions decide how much of the value an industry creates stays inside it, and how much is claimed by suppliers, buyers, substitutes, entrants and existing rivals.

That is a question about distribution, not about opportunity. It tells you where the money goes in a market. It does not tell you whether there is money in it for you, because you are not the unit being analysed. The industry is.

So founders ask the model whether to enter, and the model answers who profits if they do. The two get confused because both feel like they are about the same market. The confusion is expensive. It sends founders away from businesses that would have worked, and into fights over margin that was never going to be theirs.

Which raises the question of how much a market’s structure decides in the first place.

How much the structure explains

Strategy has an unusual piece of evidence here, because this particular question has been measured rather than argued. Take thousands of businesses, ask how much of the difference in their profitability is explained by the industry they sit in and how much by the business itself, and you get a number.

Rumelt ran it first, on Federal Trade Commission line-of-business data. Stable industry effects came out at 8 per cent of the variance in one sample and 4 in the other. Stable business-unit effects came out at 46 and 44.

Porter did not accept that. With McGahan he ran it again on a far broader sample: fourteen years of Compustat segment reports, 72,742 observations across 628 industries, every sector of the American economy except finance. Their result put industry considerably higher, at 19 per cent, against 32 for the business itself.

The two studies partition the variance differently, which is most of why they disagree, and the argument between them is still running. What survives it is the ranking. In every published specification, including the one built by the model’s own author to defend the importance of industry structure, what a business does explains more of its profitability than the industry it does it in. The margin runs between roughly two and six times. It never inverts.

There is a reason founders over-weight the smaller half anyway, and it is not carelessness. Before a venture has results, structure is the only half that can be observed. The firm-specific half has no data yet, because it has not happened. So the model gets trusted for the least interesting reason available: it is the part that is visible first.

Both datasets are made of established firms with settled industry membership, which is the population where structure should matter most. A venture that has not finished choosing its industry inherits less of it, not more.

So the model is real, and it is bounded. Structure is not noise: a fifth of the outcome is worth reading before committing years to it. But it is not the thing that decides, and a founder who treats a structural verdict as a verdict on their venture has mistaken a fifth for the whole.

Which claims are yours to move

Read the five forces as competitors on a list and you get a checklist. Read them as claims on a single unit of revenue and they become an accounting identity. Suppliers take a cut before you sell anything. Buyers negotiate part of it back. Substitutes cap what you can ask in the first place. New entrants arrive once you have proved the margin exists, and rivals compete away what is left. What survives all five is yours.

Put that way, the useful question is not how strong each claim is. It is which of them your own decisions can still move.

Two of the five cannot, at seed stage. You cannot raise a barrier to entry you have not built yet, and you cannot change how many companies are already in the market. Both are facts about the room.

The other three are not facts. They are consequences.

Supplier power is usually concentration, and concentration is a choice you made when you took the cheapest channel. Buyer power is usually switching cost, and switching cost is a design decision about where the product sits in someone’s week. Substitutes are the subtlest of the three: the alternative exists whatever you do, but doing nothing is far more attractive to a buyer with a tolerable problem than to one with an expensive one, and which of those you sell to is entirely yours.

Two of the five are facts about the room. Three are consequences of decisions you have already made.Dr. Hafiz Muhammad Ali

This is where the two numbers stop being trivia. The industry’s fifth is mostly the two you cannot move. The business’s third is mostly the three you can.

The decision it forces

So the model earns its place once, before entry, and it settles one question. Not whether the market is attractive. Whether the part of the margin you are counting on is already spoken for, and by whom.

If the answer is suppliers, or buyers, or the buyer’s option to do nothing, you have a design problem, and design problems have solutions. Change the channel. Change where the product sits. Change who you sell to.

If the answer is that margin is thin because the industry is crowded and entry is cheap for everyone, that is a structural answer, and the honest response is to accept it rather than to out-execute it. Founders lose years to that second case, because effort feels like the appropriate response to any obstacle. It is not. Some of the margin was never available, and the only decision left is whether you are willing to spend a year earning it for somebody else.

THE RULE

Three questions, one for each decision.

1. If this works, who gets paid before you do?

   Every venture has a party that takes its cut first: a platform, a marketplace, an ad network, a distributor. Name it in one sentence and put a number on it. If you cannot, you are forecasting revenue you have not established a claim to.

2. Which of your margin problems is a design decision?

   Discounting, silent churn and channel dependency feel like market conditions. They are usually consequences of where you chose to sell, who you chose to sell to, and how deeply the product sits in someone’s week. Sort them before you treat any of them as fixed.

3. What are you trying to out-execute instead of accept?

   Some structural facts do not yield to effort. If the plan requires an industry to behave differently than it has for a decade, that is not competing, it is hoping. Name the one thing you have been treating as a challenge that is actually a constraint.

Five Forces will not tell you whether to build. It will tell you who is already holding the margin, and which of their hands you put there yourself.