How to Analyse Competitive Forces in an Industry
Start from a reading that is already finished and test it line by line. Beside each of the five forces, write the fact the line rests on, where that fact came from, and the date it was true. Any line carrying only a judgement word moves to not established. Then count the answers inside each surviving line. One line can hold two. The finished note carries a count and a date.
Here is the thing worth holding on to before anything else. A reading of a field is either evidence or it is an opinion, and what separates the two is not confidence, not length and not how carefully the person worked. A reading is evidence when somebody else, handed the same fact and none of the words, would arrive at the same reading. A line carrying a fact can be checked by a stranger, laid beside the same line a year later, and corrected when it moves. A line carrying only a word can be agreed with or disagreed with, and there is nothing else anybody can do to it.
The second thing is a property of the frame rather than a fault in anybody's work. The frame allots each force exactly one line, and some forces have two answers. Where two published facts about one force point in opposite directions, writing a single word on that line throws one of them away, and nothing in the frame settles which one to keep. The honest form of that line is both answers written out and no single word at the end of it.
The third is the field most notes leave off altogether. Every fact was true on a date, and a reading is a description of a field at a moment rather than a standing property of it. The words strong and weak carry no as-of dateThe date a figure actually describes, not the date somebody wrote it down and not the date it is being read. at all. The facts underneath them move, and the words travel intact into next year. Undated words over moving facts are the specific way a correct reading turns into a wrong one without anybody making a mistake.
What is being audited, and why the audit is a separate job
A reading of a field's competitive forces arrives on the analyst's desk as five lines. The frame those five lines were written to is Michael Porter's, set out in Competitive Strategy: Techniques for Analyzing Industries and Competitors, 1980, and putting it to a field is a procedure of its own, set out under How to Apply Porter's Five Forces to an Industry. The audit begins one step after that procedure stops. The reading exists. Somebody has already done the work. The remaining job is to decide how much of it can be believed.
The job needs its own procedure because of one uncomfortable fact. Each of the five lines was supposed to rest on an observableSomething in the world that can be counted, read or timed directly, rather than something that has to be taken on somebody's word. of its own, and a reading built from written facts and a reading built from impressions come out looking exactly the same: five lines, one word each, no blanks. Nothing on the surface of a finished note says which of the two it is. The only thing that separates them is what sits underneath each line, and a finished note is precisely the document that does not show what sits underneath each line.
So the audit runs six steps, in a fixed order, and every step has a written pass and a written fail. The order matters: each fail hands the step after it something specific to do. How each of the five forces works is set out under its own heading. The audit produces no new reading of the field, only a test of the one that already exists.
What separates a reading of a field's competitive forces that is evidence from one that is an opinion?
Step one: does every line carry a fact, a source and a date?
The five lines are taken one at a time, and three things are written beside each of them, in words, rather than referred to as findable elsewhere. The fact the line rests on. Where that fact came from. The date it was true. Three fields, five lines, fifteen boxes, and the audit is decided by how many of the fifteen can actually be filled.
| Step one | The condition, written out |
|---|---|
| What to write | The fact, the source, and the date the fact was true, for each of the five lines |
| It passes if | All three fields are filled in words for that line |
| It fails if | Any one of the three is blank |
| The fail hands forward | That line moves to the not established column, and step two runs on the surviving lines only |
A line with no source is not a weak line, it is an uncheckable one, and the difference is not a matter of degree. Think of a repair quotation. A quotation carrying the price, the workshop's name and the date can be argued about, taken to another workshop, or held up next year when the same repair costs more. The same price written on the back of a scrap of paper can only be believed or not believed. Nothing else can be done with it, by anybody, ever.
The note under audit reads a single city's field for hard-bound registers built to a school specificationA written statement of exactly what a thing has to be: paper weight, number of sheets, binding, cover, size. Two different makers working to the same one produce the same article., seen from the makers' side, and its facts are those of one of them, Anjani Stationers Private Limited, an invented maker. The buying side line fills all three: one buyer at 30.00 per cent of revenue and 40.00 per cent of the money still owed, with a collection periodThe average number of days between sending an invoice and the money arriving, worked out from what is still owed against a year of sales to that customer. of 171 days against 110 days for every other account, from a customer list, as at one published year end. The input side line fills all three as well: 46.0 per cent of outside payments going to a supplier with nine known alternatives and about two weeks to replace, and 4.0 per cent going to one with a single alternative and about twenty six weeks, from the arrangements themselves, as at the same year end.
The other three lines carry a sentence each and nothing underneath it. Rivalry says competition is moderate. The line about somebody new setting up says entry is difficult. The line about what a buyer turns to instead says there are few alternatives. Each is a sentence somebody could defend in a meeting and none of the three has a fact, a source or a date attached. Three of the five lines fail step one, and they fail it before anybody has argued about whether the words are right.
| The note as it arrived | What the line said | What sat underneath it |
|---|---|---|
| Rivalry among the makers | moderate | Nothing |
| Somebody new setting up | low | Nothing |
| What a buyer turns to instead | few | Nothing |
| The buying side | significant | 30.00 per cent of revenue, 40.00 per cent of the money still owed, 171 days against 110, as at one published year end |
| The input side | high | 46.0 per cent of outside payments at nine alternatives and two weeks, 4.0 per cent at one alternative and twenty six weeks, as at the same year end |
What in the audited note is Indian, and what is not?
Three things and nothing else. Anjani Stationers Private Limited carries an Indian company form, and the audit needs nothing from that form beyond its name. The school terms that refill the order book each spring are Indian school terms. And the single city the makers work is an Indian city. The audit itself carries no jurisdiction at all: writing a fact, a source and a date beside a line, and refusing the line that has none, is the same work in any country and under any set of rules.
A line in a note reads that competition is moderate, with nothing beside it. What does step one do with it?
Step two: strike the judgement word, and does it come back?
Each surviving line is split in two. The judgement word goes on one line and the fact goes on the line underneath. The word is then covered, the fact handed to somebody who was not in the room, and that person asked what they would write. If they write the same word, the word was a summary. If they write something narrower, the word was travelling on credit.
| Step two | The condition, written out |
|---|---|
| What to write | The judgement word on one line and the fact beside it on the line below, separated |
| It passes if | A second reader, given only the fact, arrives at the same word |
| It fails if | The word carries something the fact does not |
| The fail hands forward | Keep the fact, drop the word, and let the line say what happened rather than how it felt. Step three counts what is left |
A judgement word is a summary of a fact that has not been shown, and striking it is the cheapest test in the audit. On the note under audit the input side line says the input side holds the terms. With that word covered and only the facts handed over, a second reader gets to one counterparty at 4.0 per cent of outside payments, holding a single alternative and about twenty six weeks of replacement timeThe number of weeks from the morning a counterparty stops until the business is running normally on somebody else, including finding them, fitting them and getting them working., as at the published year end. The second reader does not get to a statement about the input side in general. Another 46.0 per cent of the same payments sits with a supplier holding nine alternatives and about two weeks. The word said more than the fact, so the word goes and the fact stays.
The buying side line takes the same treatment and returns the same result in one clause: its facts describe one named buyer and its word described the buying side. Keep the facts about that buyer and drop the general word. Both surviving lines now carry facts and no words at all. Step three needs them in exactly that state.
A note says the input side holds the terms. The facts are 46.0 per cent of what goes out, holding nine alternatives at two weeks, and 4.0 per cent holding one alternative at twenty six weeks, both as at the published year end. What does step two return?
Step three: how many answers are inside each surviving force?
Now stop treating a line as a line. Take each surviving force and list every published fact bearing on it, separately, without merging them into a summary. Then count. Most of the time the count is one and the step is over in a second. Sometimes it is two, and the two point in opposite directions, and that is the case the frame has no room for.
| Step three | The condition, written out |
|---|---|
| What to write | Every published fact bearing on that force, listed one under another rather than merged |
| It passes if | The force holds one answer |
| It fails if | It holds two that point in opposite directions |
| The fail hands forward | Write both answers on the line and no single word, and carry the opposition into step four to be dated rather than resolved |
The frame gives each force one line and the field does not always supply one answer, and this is the exact point at which a careful note quietly throws away half of what it knows. On the reading under audit both survivors fail step three. The input side holds one counterparty at 46.0 per cent of outside payments with nine alternatives and about two weeks, and another at 4.0 per cent with one alternative and about twenty six weeks. Which of those two is the larger exposure is a real question with a settled answer, and that answer is worked out under The Value Network. Step three does not rank the two facts, it records that the force holds both.
The buying side holds two as well, and the two are stranger. Both are about the same buyer. The Sunrise Public School group collects at 171 days against 110 for every other account and it has stopped paying. The same group's order bookThe orders a business has taken and not yet delivered, taken together. The total tells a business how much work is standing in front of it. refills each spring without anybody at Anjani Stationers having to persuade the schools all over again. One buyer, two published facts, pointing opposite ways, on one line the frame allots a single word.
The Sunrise Public School group collects in 171 days against 110 for everyone else and has stopped paying, and the same buyer's order book refills each spring without persuasion. Before reading on, what goes on the buying side line?
Step four: what date was each fact true, and what would move it?
Every fact that survived now gets two more things written beside it: the date it was true, and one sentence naming what would change it. The second is not a forecast and it is not a guess about the field. The sentence plainly names the event that, if it happened, would make the fact wrong.
| Step four | The condition, written out |
|---|---|
| What to write | The date each surviving fact was true, and one sentence naming what would change it |
| It passes if | Every fact carries a date and something nameable would change it |
| It fails if | A fact carries no date |
| The fail hands forward | That line cannot be laid beside last year's or next year's, so step five records it as undated rather than as current |
The vintage of these particular facts is itself published, so step four bites hardest on this note. The 30.00 per cent of revenue is a total across one whole trading year. The 40.00 per cent of the money still owed is a position on one evening at the end of that year. And here is the part worth sitting with: both of those figures had been sitting on the sheet for a full year by the time the Sunrise Public School group stopped paying. Nothing was wrong when they were written down. The facts were true, the reading followed from them, and then the field moved while the words stayed exactly where they were.
The input side facts take the same treatment. The twenty six weeks holds only while one firm is the only one within reach certified on that folding and gluing machine, so a single further certification moves it. The two weeks and the nine alternatives hold while nine mills are still making that weight and finish, so a mill closing moves that one. A reading is a description of a field at a moment, and the words strong and weak carry no date at all. A doctor's note from two years ago is not wrong, and it is also not about the person sitting in the room today.
The Sunrise Public School group had already been carrying both of those figures, as at the published year end, for a full year at the point it stopped paying. What does that establish about a reading built on them?
Step five: which fact would settle what is still open, and has anybody got it?
Two kinds of line reach this step. The ones that failed step one and are sitting in the not established column, and the oppositions from step three that are still standing. For each of them, write one sentence naming the fact that would settle it. Not a topic, not an area to look into. The fact.
| Step five | The condition, written out |
|---|---|
| What to write | For each unestablished line and each standing opposition, one sentence naming the fact that would settle it |
| It passes if | That fact exists and can be got, in which case the line stops being a gap and becomes work |
| It fails if | Nobody has it |
| The fail hands forward | Step six records the question as standing, with the fact named, so next year's reader knows exactly what to look for |
Run it on the rivalry line, the line that failed step one. The fact that would settle it is how many makers the city holds and what they charge. Now check whether anybody has that. Bhavani Register Works is a promoter runRun by the person or household whose own money started the business, rather than by salaried managers answering to shareholders outside it. maker of the same registers to the same specification, making 1,50,000 registers a year, and neither what it charges nor what it pays for its paper is published anywhere. Nor does anything published say how many makers the city holds altogether. So nobody has the fact in reachable form, the step fails, and the question is written into the note as standing with the fact named.
The other two failed lines get the same one sentence each. For the line about somebody new setting up, what would settle it is what it costs to fit out a works to that specification and how long before it is delivering. For the line about what a buyer turns to instead, what would settle it is what a school actually buys when it does not buy a hard-bound register. Neither is published, so both stand. The frame contains no rule for which of the five forces wins: it produces five readings rather than a score, and adding or averaging them would need a weight nobody has published.
No published source gives the number of makers working the city, and none gives what Bhavani Register Works charges. What does step five put in the note?
Step six: what does the finished note actually say?
Everything now goes on one sheet: the lines that survived, the oppositions that stand, the standing questions with their named facts, and one date covering the facts underneath. Step six is where the audit either produces something a stranger can use next year or quietly turns back into the note it started as.
| Step six | The condition, written out |
|---|---|
| What to write | The surviving lines, the oppositions, the standing questions with their named facts, and one date |
| It passes if | The note says how many of the five were established, carries a date, records both answers wherever a force held two, and names each standing question with the fact that would settle it |
| It fails if | The note produces one word for the field |
| The fail hands forward | Nothing, and that is the cost. The word is what gets quoted a year from now, and none of the rest of this will be attached to it when it travels |
| The note as it left | What it now says |
|---|---|
| The buying side | Two answers, both dated: one buyer at 30.00 per cent of revenue and 40.00 per cent of the money still owed, collecting at 171 days against 110 and no longer paying; and the same buyer's order book refilling each spring without persuasion |
| The input side | Two answers, both dated: 46.0 per cent of outside payments at nine alternatives and about two weeks; 4.0 per cent at one alternative and about twenty six weeks |
| Rivalry | Not established. What would settle it: how many makers the city holds and what they charge. Nobody has it |
| Somebody new setting up | Not established. What would settle it: what it costs to fit out a works to the specification and how long before it delivers. Nobody has it |
| What a buyer turns to instead | Not established. What would settle it: what a school buys when it does not buy a hard-bound register. Nobody has it |
| The whole note | Two of five established, both of them holding two answers, on facts as at one published year end, with three questions standing and the fact that would settle each one named |
Notice what is not there: a word for the field. No summary, no rating, no verdict, and nothing that could be lifted out of the note and repeated in a corridor. A count and a date can be lifted out, and both carry their own limits with them wherever they go. Nobody has established a threshold at which a count of established lines becomes enough or a fact becomes too old.
Why does step six refuse to let the note end in a single word for the field?
What does this frame not settle?
Three things, and none of them is a fault in the frame. The frame allots each force one line, so it does not settle a force that has two answers. The frame carries no date of its own, so it does not settle when a reading stops being true. And nothing in the frame supplies a weight, so it does not settle how the five combine.
All three are reasons to write a reading down more fully rather than reasons to distrust it. A reading with its facts, its dates and its oppositions attached is worth more next year than a confident word was worth this year, and the whole six step audit exists to produce that fuller reading. The frame is doing a frame's job, naming which five questions to ask. The frame never promised to say whether anybody had answered them.
Before reading on, why does a five forces note so often come back with all five lines filled in?
The note where all five came back filled in
The filled-in note comes from somebody working hard rather than from somebody cutting a corner, and the hard work is exactly the problem. The frame arrives as five lines. Five lines look like five questions, and a question put in front of a careful person gets answered. When a fact is to hand, the answer is written in facts. When no fact is to hand, the answer is still written, in the only currency left available, a word: moderate, low, few, significant.
The finished note has no blanks in it. The note reads as complete, it reads as balanced, and three fifths of it rests on nothing at all. Name what actually happened without blaming anybody for it. The fault is a property of the layout rather than a lapse in the person holding the pen. The analyst is trying to avoid a blank line, and the blank line is the finding.
The two versions of the same reading, side by side
The filled version returns five words and reads as a complete picture of a field. The audited version returns two lines carrying facts, both of which hold two answers pointing opposite ways, three lines reading not established with the fact that would settle each one named, and a date. The second is shorter and says considerably more.
The first cost: it cannot be checked
A word has nothing underneath it for anybody to disagree with. Hand the filled note to a colleague and the only conversation available is whether they would have written the same five words, a conversation about two people rather than about a field.
The second cost, and this is the one that compounds
The filled note cannot be laid beside next year's. A collection period and a replacement time are the kind of thing that moves and can be seen moving, so two of the audited lines can be put next to the same two lines twelve months later and the movement read straight off. Five words laid beside five words show only whether somebody's mood changed.
And the fix is not more research, and it is not a longer note. The filled version already contained every fact the audited version uses. Missing from it were the three fields, the count inside each force, and the date. A line that says not established can be filled in next year. A line that says moderate cannot be corrected. Nobody will remember there was nothing under it.
Two notes on the same field are laid beside their own versions twelve months later. One holds five confident words, the other two fact-bearing lines with dates. What can be read off?
How somebody who inherited the note actually uses this
The analyst who did not write the note has to decide how much of it to believe, usually with somebody waiting for an answer. The card below carries the whole audit in order, showing on each row what lets a line through, what stops it, and what this particular reading returned. The order is the argument: step three only makes sense on lines that survived step one, and step five only has work to do because steps one and three left something open.
The card is worth keeping for one reason. A note written this way can be laid beside next year's version of itself and the differences read straight off, and the whole audit was only ever for that comparison.
The audit tests a reading of a field's competitive forces that somebody else has already finished. Producing such a reading in the first place is a different procedure entirely, covered separately under How to Apply Porter's Five Forces to an Industry.
Each of the five forces has notes of its own. Look for Buyer Power: When Customers Set the Terms. Then Supplier Power: When Inputs Set the Terms. Then Substitutes: The Competition That Is Not in the Industry. Rivalry between the makers already standing in a field sits under the heading Competitive Rivalry: How Intensity Shapes Industry Returns, and how easily somebody new could set up is treated in The Sources of Competitive Advantage.
Sorting evidence by where it came from, before any work starts, belongs to a separate treatment written under the heading Company Analysis vs Industry Analysis. Measuring how concentrated a field is, and what happens to that measure when nobody agrees where the field ends, belongs to another one, written up as Herfindahl-Hirschman Index. How a field changes shape across the years belongs to a third, The Industry Life Cycle: Emergence to Decline, and step four here is about when a single fact was true rather than about how fields evolve. Which of two counterparties is the heavier exposure, the one taking much of what goes out or the one nobody can replace, is settled in a fourth, The Value Network.
The audit produces a count with a date. A count with a date puts no price on anything and reaches no verdict about the field.
What was consulted, and what date does each figure carry?
| What is being named | The document | Where to read it |
|---|---|---|
| The frame the audited note was written to | Michael Porter, Competitive Strategy: Techniques for Analyzing Industries and Competitors, 1980, the origin of the frame and the source of no figure | The publisher's catalogue listing for the 1980 edition |
| The buying side facts audited at steps one, three and four: 30.00 per cent of revenue across one whole trading year, 40.00 per cent of the money still owed as at the close of that year, and collection at 171 days against 110 on that same year | The earlier notes on customer concentration and on what a name is worth, covering the Sunrise Public School group and the year that passed before it stopped paying | These notes, Customers and Brands, on finmaverick.com |
| The input side facts audited at the same three steps, all as at the close of the same trading year: 46.0 per cent of outside payments with nine alternatives and about two weeks, and 4.0 per cent with one alternative and about twenty six weeks | The earlier notes on the network of counterparties, which publish the six shares of outside payments with their alternatives and replacement times | These notes, Business Models, on finmaverick.com |
| Bhavani Register Works, promoter run, 1,50,000 registers a year to the same specification, with what it charges and what it pays for paper published nowhere, which is what leaves the rivalry line standing open | The earlier notes introducing that second maker | These notes, Competitive Advantage and Moats, on finmaverick.com |
Anjani Stationers Private Limited, Bhavani Register Works and the Sunrise Public School group are invented.
Educational material. Not advice on any investment, tax, budget or market position.
