Substitutes: The Competition That Is Not in the Industry

There is a particular kind of competition that never turns up where anybody looks for it. Nobody finds it on the list of names an analyst builds, and nobody finds it in the year's accounts. No quotation was ever asked for, so none was undercut. Such competition is the reason a business can look untouched for years by every measure available and still be losing the argument. The evidence for it lives nowhere near the business, and that distance makes it the hardest of the five forces to establish.

Substitutes are one of the five forces Michael Porter set out in Competitive Strategy, 1980. How hard the sellers already in a trade fight each other is a separate force. Whether some businesses face fewer alternatives than others is a separate question again, settled from the buyer's side. Substitution is the competition that arrives from outside the trade altogether, and it refuses to show up in two places: the competitor list and the accounts.

One definition and two businesses carry the whole argument. The first is Anjani Stationers Private Limited, an invented maker of hard-bound registers for schools. Its contribution marginWhat is left out of every rupee of sales once the costs that move with the sale have been taken off. It is a share, so it can be read across two years of different sizes. was 42.75 per cent in year one and 42.78 per cent in year two, on revenue of Rs 2,40,00,000/- and then Rs 2,70,00,000/-. The second is Setu Bazaar, an invented online marketplace across which goods worth Rs 5,00,00,00,000/- passed in its published year.

And the argument rests on one ruling about what a set of accounts can establish. Ruling a competitor in needs three facts that sit in no statement of profit and loss anywhere: the count of sellers in the field, the ease with which one more of them could set up shop, and the other routes standing open to a buyer. So a set of accounts can rule a competitor out and can never rule one in. Substitution is that ruling at its sharpest. A rival at least leaves a quotation somebody lost, and a substitute takes the buyer before there is anything to lose.

What is a substitute, and what makes it different from a rival?

Begin with the buyer rather than with a definition. The definition is only useful after one small exercise. Take a customer of Anjani Stationers and write down what that customer was trying to achieve. One sentence. And write it without using the name of the thing being sold.

A school is not trying to acquire a hard-bound register. A school is trying to have a record of something that lasts the year, can be written in by hand, can be signed by whoever is responsible for it, and can be produced again a term later when somebody asks. The purpose is that record. The register is one way of meeting it, and the word one is doing all the work in that sentence.

The exercise has already moved the boundary. The moment a purpose is written down without the product inside it, the set of people competing for that purpose gets longer, and it stops being a set of people who resemble the seller. A rival competes for the order and a substitute removes the need for the order. A rival is defined by what it sells; a substitute is defined by what the buyer was trying to achieve. Two circles drawn around the same buyer, and the second is always the larger of the two.

Take it off the street for a moment. Two bus operators run the same route between the same two points. The two operators are rivals in the plainest sense: same journey, same passengers, same fare, and each can take a passenger off the other on any given morning. Now a neighbour who used to catch that bus starts walking to work instead. The neighbour runs no vehicle, keeps no timetable, charges nobody, and could not appear on a list of bus operators however carefully that list were compiled. And a fare has gone from both businesses.

Why does a substitute never appear on a list of competitors?

Watch somebody build a competitor list and the answer becomes obvious rather than clever. The method is always the same. Name a trade, then find everybody in it. Stationery makers. Marketplaces. Paper mills. Somebody chose that category, and the choice decided who counts and who is never considered at all.

The list is built from the trade and the substitute is built from the job, so a complete list and a complete picture are two different things. The natural reaction is to call the list careless, and the reaction is worth sitting with. The list is not careless. A list of every stationery maker in a district, checked name by name, is a good list and a correct list. Such a list is complete against the question it was asked, and it simply was not asked the buyer's question.

Anything meeting the same purpose from outside the category fails the test by construction. Not by oversight, not because somebody forgot it, but because the test itself was membership of a trade, and the thing in question is not a member of that trade and never was. No amount of diligence inside the boundary reaches outside it.

A household picture makes it concrete. Parents comparing three tuition centres for a child have a complete list of tuition centres in the neighbourhood, with fees, timings and results all laid out. The sheet leaves out the elder cousin at home who would teach the same syllabus for nothing. The cousin has no fee to compare, runs no centre, advertises nowhere, and could take the whole decision away.

Where the boundary of a market should be drawn in the first place, and what a badly drawn one does to a measure of concentration, is taken up alongside the Herfindahl-Hirschman Index. Here it is enough to notice that somebody drew a boundary, and that whoever drew it decided what could appear.

Why does a substitute not show up in a set of accounts either?

The list is the half most people see. The accounts are the half that catches them out. The reasoning is short and the consequence is large, so it is worth going slowly.

Ask what each kind of competition leaves behind. A rival who undercuts leaves a trail whether anybody wants one or not. Somebody asked for a quotationA priced offer a seller puts in writing before any order exists, so that a buyer can compare it with somebody else's. Losing one is an event that can be dated and recorded. and it was lost to a lower number. A charge that had stood for three years had to come down to keep an account. Neither of those is pleasant to look at, and both of them are visible. Both can be dated, counted and argued about.

A substitute leaves neither. The buyer who decided to get the job done another way never asked for a quotation at all. Nothing was quoted, so nothing was undercut. Nothing was ever in play, so nothing was lost. A substitute takes the order before the order exists, and that is why it is the one force with no invoice attached to it.

The uncomfortable consequence follows directly and people skip it, so say it out loud. A business can be losing to a substitute for years while every line of its accounts looks entirely ordinary. Revenue up. Margin steady. No customer complaints on file about price. The people who went elsewhere were never customers and had nothing to complain about.

The street version of this is a tea stall that never notices the office two doors down installed a kettle. Nobody came in to argue about the price of tea. Nobody switched to the stall across the road. Twenty regulars simply stopped walking through the door, one at a time, over a couple of months, and the stall's own record of the day contains nothing whatsoever about any of it.

Where does a substitute land first, on the charge or on the volume?

On the charge, and it gets there without taking a single order. The pressure on the charge is the part of the force that operates while nothing at all appears to be happening.

A substitute sets a ceiling. A substitute does not have to win anybody over to set one. The substitute only has to exist and be known. Past some charge the buyer takes the other route instead, and both sides of the table can feel roughly where that point sits without either of them naming it. A seller who could raise a charge and does not, out of a rough sense of where the buyer would stop, has already met the force.

A substitute can cost a business the rise it never attempted, and a rise never attempted appears nowhere. There is no line for it. There is no meeting where it was decided against, or at least no minute of one. The order book still fills, the customers still come back, and the ceiling sits quietly above the charge doing its work.

Nothing published says what any buyer would tolerate, for either business, so the ceiling carries no figure. An invented number would look researched because it sat beside real ones, and it would be quoted later without the warning. A quantity nobody has measured is drawn honestly only with no scale on either axis, and the ceiling below is drawn that way.

A lending library shows the shape of it. The library has not lost a member in years. Every subscription renews. And everybody in the neighbourhood knows what the second-hand shop charges for a book, so the person running the library cannot raise the subscription by very much. The moment the subscription drifts past that comparison, the arithmetic changes in the member's head. Nobody has left. The ceiling is still real.

What do Anjani Stationers' two published years actually settle?

Put the pair on the table and read it slowly. The reading has two halves, and almost everybody drops the second one.

Divide each contribution by its own revenue and the two margins come back out. Publishing the components rather than the ratios is what makes that check possible. Rs 1,02,60,000/- over Rs 2,40,00,000/- is 42.75 per cent. Rs 1,15,50,000/- over Rs 2,70,00,000/- is 42.78 per cent. The margin moved by 0.03 of a point, and it moved upward. Revenue was 12.50 per cent higher in the second year, at Rs 2,70,00,000/- against Rs 2,40,00,000/-.

Now the reading. Across those twenty four months the price side did not give way and the volume side did not shrink. Whatever else was going on, nothing outside the trade took anything off either of them inside that window. The pair acquits and convicts nothing, and the acquittal covers twenty four months and not a day more.

The last clause is not a caveat bolted onto the finding but half the finding itself. Two observations of one business describe the span between those two observations, and a substitute that arrived in the twenty fifth month would leave both of those readings exactly where they are.

Two comforts a reader will reach for now, and both have to be taken away. The first is that the Sunrise Public School group has bought from Anjani Stationers for eleven years and accounted for 30.00 per cent of the second year. Eleven years is a fact about one account, and a very good one. A long account is not a fact about whether another way of keeping the record exists. The second is that the order bookThe work a business has already been asked for and has not yet delivered. A full one says what has been promised; it says nothing about what was never asked for. refills each spring on the school calendar, without anybody having to persuade the schools again. A refilling order book is a fact about a calendar. Neither can be turned into evidence about the world outside the trade, however firmly they are held.

What is the substitute for a marketplace like Setu Bazaar?

Nothing published says what a substitute for a register would be, so Anjani Stationers gives a clean acquittal and no worked substitute. Setu Bazaar gives the opposite, and so the worked example is Setu Bazaar's.

Here is its published year. Setu Bazaar has 50,000 buyers and 2,000 merchants. Goods worth Rs 5,00,00,00,000/- crossed it in the year. The figure is the gross merchandise valueThe value of everything traded across a marketplace in a period. It is what buyers spent, not what the marketplace earned, and the two are usually very far apart. of the year rather than anything Setu Bazaar earned. Setu Bazaar kept 4.00 per cent of that, Rs 20,00,00,000/- of revenue, and the remaining Rs 4,80,00,00,000/- reached the sellers.

Now the alternative, and it is published as an ordinary feature of marketplaces rather than as anybody's threat. Two sides that Setu Bazaar first brought together can settle the next trade between themselves. The name for it is disintermediationWhoever made an introduction being cut out of everything that follows it, because the two sides go on dealing on their own and stop paying for a connection they already hold.. Nothing stops them. The goods still move, the money still changes hands, and Setu Bazaar is simply not in the path any more.

The alternative to a marketplace is not another marketplace, it is no marketplace. The sentence is the whole force in nine words, and the contrast that proves it is sitting right beside it. A merchant that sells through three marketplaces at the same time is multi-homingWorking through several rival services in parallel rather than settling on a single one, which leaves any of them easier to abandon than it would otherwise be., and multi-homing is trade moving to a rival. Everything in that picture is still a marketplace and every name in it would be on a competitor list. The direct deal is nobody taking the trade at all.

Multi-homing and the direct deal stay apart. Both arrive in the same sentence where they are published, both loosen a marketplace's position, and they belong to two different forces. Multi-homing is somebody else winning. The direct deal is nobody winning and the trade leaving the category.

How much of Setu Bazaar's year goes direct is published nowhere. The panel below lets a share be set and shows the effect on the published year, and every share set there is a setting rather than a fact about Setu Bazaar.

What has to be observed before saying a substitute is doing anything?

Three things, and they are worth learning as a set because a claim missing any one of them is not yet a claim.

The first is whether another way of getting the job done exists at all, described from the buyer's purpose rather than from the seller's own product. The second is what it would cost that buyer to move to it, counted in money, in effort and in habit. Moving cost is a subject of its own, taken up under Switching Costs: Why Customers Stay Even When They Could Leave. The third is whether that other way is getting cheaper or better. Direction is a different thing from level, and direction is the one of the three that decides whether a ceiling is coming down.

Every one of the three is a fact about somebody who is neither a customer of the business nor a rival to it, so every one of them has to be gone out and asked for. Not one of them can be derived, inferred or estimated from a set of accounts, however many years of them are stacked up. The accounts record what was bought from this business, and these three questions are about people who bought nothing from it.

Now the compulsory declaration. For Anjani Stationers' registers, not one of the three is published. The alternative is not named, no cost is put on moving to it, and nothing is said about any direction. For Setu Bazaar, exactly one of the three is published: nothing stops a buyer and a merchant dealing directly once they have met. The other two are as blank there as they are for the registers.

Why can a set of accounts rule a substitute out and never rule one in?

The ruling is worth carrying in one form. A movement that did not happen cannot have been caused by anything. So a charge that held cannot have been pushed down, and a margin that sat at 42.75 per cent and then at 42.78 per cent cannot have been the site of anything taking the price side away. The finding is an acquittal, it is real, and it is worth having.

The other direction fails, and on this force it fails for a reason sharper than on any of the other four. To convict, the buyers who are not in the accounts would have to be counted, and the whole point is that they did not buy. The people who would prove the case are precisely the people who left no trace, and no depth of detail about the ones who stayed can reach them.

The gate opens one way, and on this force the closed side is closed harder than on any of the others. A rival can at least be caught in the accounts indirectly, through a quotation lost or a charge that had to move. A substitute cannot be caught there at all, in any form, ever, no matter how the statements are cut.

The consequence is a rule about where to put the effort. A claim that a substitute is hurting a business is a claim about the world outside that business. Such a claim has to be supported with facts from out there, gathered by somebody who went and looked, or it has to be dropped. There is no third option in which more work on the accounts eventually produces it.

How is all this put to a claim somebody hands over?