Competitor Mapping: Drawing the Field Before Judging Position

A competitor map is a boundary decision before it is a list of names. Who counts depends on whether the line is drawn round the trade or round what the buyer was trying to achieve. The line gets drawn, every cell gets filled with a named fact or with the words not established, and there the work stops. The empty cells are the result, and they are drawn the same size as the full ones.

Somebody asks who this business competes with. What was the second question?

Start with the request rather than with a definition. The request is where the damage happens, and it happens in the first four seconds. Anjani Stationers Private Limited, an invented maker, turns paper into hard-bound school registers to a school specificationAn agreed list of what a made thing has to be: binding, paper weight, ruling, cover size. Two makers working to one of these turn out goods a buyer treats as interchangeable. and sells them to schools in one city. Somebody hands an analyst that description and asks a perfectly ordinary question. Who does this business compete with?

Almost everybody starts listing. Another register maker. Somebody bigger. Somebody cheaper. And within a minute there is a list, and the list feels like progress because it has names on it and names look like findings. Nobody notices that a second question was asked at the same moment and was answered silently, by whoever started listing first. Competes at what?

Take the two honest readings of that request and hold them side by side. On the first reading, a competitor is another maker of hard-bound registers to the same specification, selling to the same schools. On the second, a competitor is anything at all a school could buy instead in order to keep a record that lasts the year. Both readings are defensible. Both would be recognised as reasonable by anybody in the trade. And they produce different sets of names over exactly the same purchases.

Sit with that for a sentence before it gets resolved. The discomfort is the useful part. If two competent people can draw two different fields round one business and both be right, then the drawing is not a discovery. The drawing is a decision, and somebody has to make it out loud.

So what actually decides who is inside the line and who is outside it?

The rule that settles the question is two lines long. A rival is defined by what it sells. A substitute is defined by what the buyer was trying to achieve. A rival and a substitute are not two competing definitions of one word. The two definitions draw two different circles, and the circle round the purpose is always the larger. Everything sold in the trade was also bought for the purpose. Plenty bought for the purpose was never sold by the trade.

The geometry has a consequence worth stating plainly, and the whole exercise turns on it. Drawn on the trade, the line encloses one set of buying; drawn on the purpose, it encloses a bigger set of exactly the same buying, and not one buyer bought differently under either reading. Nothing moved. The line moved.

The same thing happens in a shopping centre with ten shops under one roof on a Saturday. Asking who competes with the shoe shop gets one answer if the question meant the other shoe shop on the floor above, and a completely different answer if it meant everybody the same household walks past with the same money in the same pocket and a fixed number of hours. The households did not change their plans between the two questions. Somebody drew a different line round them.

What was the buyer actually trying to achieve, in the buyer's own words?

If the purpose decides how wide the outer circle may honestly go, then the purpose has to be taken in the buyer's own terms and quoted rather than tidied up. A paraphrase drifts toward the trade already in mind. The drift is how a wide boundary quietly becomes a narrow one again. For this case the purpose is published, and it carries four conditions. The school wants, in its own words, a record of something that lasts the year. Second condition, unchanged: it can be written in by hand. Third: it can be signed by whoever is responsible for it. Fourth: it can be produced again a term later when somebody asks.

Read that slowly and count the conditions rather than the words. Four of them. A record that survives a year. Something a person can write into by hand. Something a responsible person can sign. Something retrievable a term later on request. Nowhere in those four conditions does the word register appear, and nowhere does paper appear either. The absence of both words is precisely why the outer circle takes in things the trade never sold.

Something did not happen while those conditions were set out, and it is worth marking. Nobody found a new buyer. Nobody discovered a purchase that had been hiding. The set of things capable of meeting all four conditions simply got larger the moment the conditions were written down instead of assumed.

How to Build a Competitor Map: what does the procedure actually run through?

Six steps, in order, and every one of them runnable on a fact somebody actually holds. The order matters more than any single step, and step one matters most of all.

Why step one comes first is worth stating. Leaving it out costs more than any other slip available. Unless what is sold and to whom is written down before the naming starts, every later step answers that question by accident, in whatever direction the first name to occur happened to point. The boundary gets set either way; the only choice is whether it gets set on purpose or by the order in which names happened to occur.

And step six is the part most procedures lack. A procedure that runs to the end and always produces a filled sheet is a prompt: it dictates what to write, and having something to write at the end of it is certain. A procedure with a stopping rule that permits the words not established is a test. Such a procedure can come back empty. The distinction between a prompt and a test is settled separately, and applied here. A run whose steps are allowed to return nothing is testing something; a run that is certain to return an answer at every step is testing nothing, and the two are indistinguishable from the outside until one of them comes back empty.

One caution about those two words before they get used forty times. Not established records an outcome: somebody set a condition, went and searched for a fact meeting it, and came back with none. Not established is not a hedge, it is not a polite way of saying probably, and it is not a shrug. The phrase records a result, and it carries a date and a place where the looking happened.

What does the ladder look like when it is built upward from the invoices?

Step three said to start from the invoices, and this is why. The served setThe buyers who actually bought during the year, taken off the invoices issued rather than off any view about who might have bought. is the one cell in the whole exercise that is exact, and not an estimate of anything. The served set is a stack of documents. For this case it is 36 accounts and Rs 2,70,00,000/-, and it is the bottom rung of the ladder below.

Now climb, one rung at a time, and write into each cell whatever is published for it and nothing else.

Count the ladder rather than dividing it, and there are three plain facts to write down. Of five rungs, one carries money. One carries a floor in registers and no money whatever. Three of five carry nothing at all. The reader arriving at this expects the ladder to fill as it widens, and it empties instead. A wider line does not bring more evidence with it, and generally brings less.

The emptying is not an accident of this particular trade. The invoices in hand are the narrowest cell on the ladder and the only exact one, and every rung outward is a step further from anything anybody ever wrote down. The empty bands are drawn at the same height as the filled ones, and that choice is doing real work: a band squeezed, greyed or left off reads as an unfinished piece of work, and a band drawn at full height reads as what it is, a finding about the evidence.

Why does rung one carry a floor in registers and no money at all?

Because there is a reason rather than a shrug behind it, and the reason is worth stating in full. There is a second maker turning out registers to that same school specification. The second maker is Bhavani Register Works, invented for teaching, promoter runDirected by the people who set the business up, rather than by managers brought in from outside to run it for them., turning out 1,50,000 registers a year. Three facts are the whole of the published record about it.

Its charge is published nowhere. Its paper cost is published nowhere. The cost of running its worksThe premises where the making happens: the machines, the floor they stand on and the people running them, counted together as one operating unit. over a year is published nowhere. So the rung carries a unit floor and an empty money cell, and the alternative is worse than it looks. Fill that cell with a trade averageA figure taken from what businesses of a similar size and shape are generally believed to do, rather than from anything this particular business published. and the result is a line nobody can ever check. The average came from a judgement made in a room, and the judgement was never written down beside it. A cell reading not established can be filled the moment somebody gets the fact, and a cell filled with an average can never be corrected at all.

What does the map these notes can actually draw look like?

Three rows. The evidence supports no more than that. Row one is Anjani Stationers at 2,50,000 registers a year. Row two is Bhavani Register Works at 1,50,000 registers a year, to the same specification. Row three is every other maker, whose count is published nowhere, drawn with question marks and with no right-hand end at all.

The third row is the one that decides what the drawing can be used for. The third row has no edge, so there is nothing there to divide by. Take the edge off that bottom row and a share, a total and a ranking all become unbuildable from the two rows standing over it, whatever anybody would like to do with them. The bar on dividing is not a rule imposed on the drawing from outside. The missing edge is a property of the drawing and visible in it, and the row is drawn running off the sheet instead of being tidied into a box for exactly that reason.

Compare that with how the same field usually appears on a slide: three neat bars in a chart with a title reading share of the trade. The chart contains one more fact than the drawing does, and that one fact is the fact nobody has.

Two outputs sit a paragraph apart and invite a sum. Why not take it?

Two output figures now sit in adjacent rows, and the hand reaches for the addition without asking anybody's permission. Nearly every reader does it. The addition feels like the obvious next move, and the arithmetic itself is trivial and would be correct as arithmetic.

The sum goes nowhere, and the reason is short. Nothing in these notes states how many makers the city holds. In hand are two makers' outputs. A total would need a count of the makers before it could mean anything, and a count is a different kind of knowledge entirely. Counting the sellers in a field and knowing the output of two of them are two separate facts, and only the first of them supports a share. A sum of the two that happen to be known is not a small understatement of the trade. The sum is a figure with nothing to be a figure of.

The positive form of that is more useful than the prohibition. Held: two named outputs, each worth writing down beside the route it came down. Not held: an edge. Both of those statements belong on the drawing, and the second one belongs there in the same weight of ink as the first.

What goes in each cell, and why is a name not a reading?

The cleanest pair in the exercise is worth memorising in this shape. A rival existing is a fact about the trade. How the two behave towards each other is a reading of rivalry. The two claims are not the same size, and they do not rest on the same kind of evidence.

A reading of rivalry needs something underneath it before it can be written at all: a price move, a discount somebody gave, a term somebody changed, a capacity somebody added. Any one of those four would do. On this field, none of the four is published anywhere. So the rivalry line is recorded as not established, and the fact that a second maker exists goes into its own cell where it belongs.

The cell shape, stated once and used everywhere on the drawing: a name, a fact, and the route that fact came down, or else the words not established. Three things or one phrase. Nothing in between, and no single adjective standing in for the lot.

What happens when the field is run in a fixed order and counted?

Running a field line by line in a fixed order is an old idea and it is worth attributing. Michael E. Porter set out five competitive forces in Competitive Strategy in 1980, and the frame is named here for the order in which a field is run through and for nothing else. The frame supplies no figure, no line round the field and no verdict about anybody in it. The judgement of rivalry itself, and the reading of a field's structure, belong to a separate treatment, routed in the table below.

Run over this field, a reading enters a line only where a published fact can be pointed at to hold it up. Two of the five lines survive that.

Now the part that gets skipped. Nothing is added, nothing is averaged and nothing is turned into a single word for the field. Two of five stand. Three go down as not established. The count itself, and nothing beyond it, is the verdict. There is no score underneath it and no summary adjective sitting on top of it. Notice too what each surviving reading is actually about: one names a single buyer and the other names a single supplier, so neither one will carry a sentence beginning schools generally or suppliers generally, however much a reader would like it to.

The two that survived are worth a closer look. Their shape shows something the count on its own does not. Each rests on facts about one identified party. On the buying side those facts are a share of the year's revenue, a larger share again of the money still owedSales already invoiced and not yet paid for. It sits as an amount due from named buyers rather than as cash the business can spend. at the year end, and a stretch of collection daysThe average number of days between issuing an invoice and the money actually arriving, worked out over a whole year of sales. far longer than anybody else takes. On the input side they are a share of outside payments, a count of known alternatives and a replacement period. Every one of those is a fact about a counterparty, and not one is a fact about the field.

And the sharpest thing the count says is about other runs rather than about this one. A run over the same field that comes back with five readings and no blanks has invented three of them. The facts that would settle those three lines do not exist to be found.

Three lines came back empty. So what actually gets written in them?

Three lines, published in these notes and turned here on the drawing's own cells. First, write the question down. Then write what would settle it either way. Then write that nobody has it. Three sentences per empty cell, and the cell stops being a hole and starts being an instruction.

Take the rivalry line as the worked case. The question: do the two named makers move on price, terms or capacity in response to one another? The evidence that would settle it: a price change by one followed by a price change by the other, a discount recorded, a term altered, or capacity added by either. Who has it: nobody, in anything published anywhere in these notes.

Now the asymmetry that turns this from a preference into an argument, and it is the whole case for doing it. A line reading not established can be filled the moment somebody gets the fact, and a line filled with an average can never be corrected. Nobody will remember it was invented. The asymmetry is not a difference of tone. The difference lies in what a later reader is able to do. Three honest blanks are three telephone calls somebody can make. Three quietly filled cells are three permanent holes that look like floor.

So when is a drawing of a field finished, and what does a finished one show?

A drawing is finished at the point where no cell is left without either a named fact in it or the words not established, and not a pass before that, and not a pass after it either. There is no further pass that improves it. The remaining cells are empty for a reason that more effort cannot touch.

A finished drawing carries three things and no more. Who is in the field, on a line somebody wrote down. Then what is known about each of them, and by what route each of those facts came down. And where nobody knows anything at all. A finished drawing does not say who is winning, it does not rank anybody, and it puts no worth on anything. None of those three is a question a drawing of a field was ever able to answer.

One separation before the end. Work of this shape sits close to a job that looks similar and is not. Drawing a field in order to say who is in it and what is known about each one is a question about the trade. Choosing a handful of businesses in order to set a price beside them is a question about a valuation. The two start in different places, they are judged by different tests, and the second is covered separately.

The claim this guide has been building the whole way down is this. The drawing that can safely be handed to somebody else is the one where the empty cells are the same size as the full ones. Only that version can be improved by a later reader rather than merely inherited.