Innovation: The Types and Which Ones Threaten Incumbents

A change in method is worth what its position makes it worth. In one works where three stages run one after another, lifting the middle stage by 25 an hour would move yearly ability not at all. Lifting the last stage by the same 25 would move it by 1,00,000 registers. So innovation splits by what a change touches and who it threatens, never by how large it looks.

One works, two lifts of the same size, and one of them is worth nothing. How?

Before any definition, watch a single works. Anjani Stationers Private Limited, invented for these notes and trading nowhere, makes hard-bound registers on a line of three stages that run one after another. The line cuts 150 registers an hour, it prints 125, and it binds 100. Nothing else about the building matters yet.

The slowest stage sets the rate for the whole line, so the works makes 100 an hour and the two faster stages spend part of their time waiting. Two lines across eight hours across 250 working days give 4,000 line-hoursOne hour of one production line being available to run. Two lines available for the same eight hours is sixteen line-hours, and the count is of availability rather than of output., and 4,000 line-hours at 100 an hour is a rated capacityWhat a works would turn out in a year if it ran at its own rate for every hour it counts as available. It is a ceiling rather than a plan, and how it is built is covered separately. of 4,00,000 registers a year. The published year turned out 2,50,000, and of the paper that went in, 70.42 per cent came back out as finished registers. The share that came back out is that works’s yieldThe share of the material that started the process and ended up inside something saleable. It is a second reading beside the rate, and it answers a completely different question..

Now the part worth slowing down for. The same works is asked what would happen if one stage were lifted, and four rows follow, worked in full separately. Not one of the four is an event. The business bought nothing, lifted nothing and reached nothing; these are four demonstrations on paper against one published works whose rated capacity is 4,00,000 registers and whose line runs at 100 an hour. Every row below is written in the conditional for that reason.

Read the last column and then read it again. The same rupee of spending would buy 1,00,000 registers of rated capacity at binding and exactly nothing at printing, and nothing about the two machines explains that difference. The difference is established where the line itself is worked, and everything that follows rests on it.

The harder half is the one people skip. The return on a faster press at printing is not a poor return that somebody might defend on other grounds. The return is exactly nothing, and the cost is worse than nothing. The press would add to the Rs 74,00,000/- of standing costCost whose total stays where it is whether the works makes one register or four lakh of them. Rent, salaries and insurance sit here, and what puts a cost in this group is covered separately. the works already carries, and the same 2,50,000 registers would then have to carry more of it. The operating resultWhat is left in a year once the costs of running the business are taken off the money it brought in, before anything to do with borrowing or tax. It is a line on a published account rather than a judgement about anybody. would fall at unchanged volume.

Three people sit at a table passing a job down the line. The first two work quickly, the third works slowly, and the pile of half-finished work sits in front of the third. A better pen for the first one buys a bigger pile. Everybody at that table can see it. Somehow, on a proposal with a supplier quotation attached, the same thing becomes invisible.

So what does the whole of this rest on?

An innovation's return is not a property of the innovation, and everything else here follows from that one sentence. The return is a property of where the change lands and of when it gets paid for. Two changes can be identical in kind, identical in size, made in the same building in the same year by the same people, and one of them moves the year while the other moves nothing. Nothing about the change itself tells which of the two it is.

A line of stages running one after another makes what its slowest stage makes, and that is arithmetic rather than an opinion about how well anybody is managing. Every rupee spent anywhere except the slowest stage therefore buys waiting time that was already sitting there. Nobody has to behave badly for this to happen. Waiting appears whenever the stages are not all the same speed, and the stages never are.

So a change carries a position before it carries a size, and the position decides everything the size later gets credited with. Changes therefore split by what they touch rather than by how large they look. A change to what is sold and a change to how it is made are two different positions, paid for out of two different things, and threatening two different sets of people.

And the threat runs the opposite way to the attention. A business already established in a field watches the changes it can see: the large ones, the ones its own trade talks about, the ones that arrive in the same shape as its own work. The change that unseats it is the one that does not look like its work at all, so nobody in the building is assigned to watch it. The size of a change and the attention it attracts move together. The size of a change and the threat it carries do not.

So what actually separates one change in method from another?

A reader probably arrives expecting a ladder: small improvements at the bottom, big ones in the middle, the field-changing ones at the top. A ladder is a natural instrument and the wrong one, and the works above is why. A ladder ranks by size, and the two lifts just worked are the same size, in the same building, in the same year, with returns of 1,00,000 registers and nought. A ladder puts them on the same rung.

Two questions replace the ladder. The first is what the change touches: what is sold, or how it is made. The second is who it threatens: the businesses already established in the field, or somebody else, or nobody at all. Neither question is about how new the change is, how much it cost or who did it first.

Two questions produce four cells, and a ladder produces a ranking. The difference matters more than it sounds. A ranking invites somebody to spend at the top of it, and the top of a ranking by size is the stage that was never setting the rate. Four cells invite a different move: locate the change's cell, then ask the cell's own question. Asking what the change touches shows where the return would have to come from. Who it threatens shows who will notice in time.

Notice what the drawing refuses to do. Both works lifts land in the same cell, and their returns are 1,00,000 registers and nothing. So the cell narrows the question and does not answer it. Anybody who wanted a grid that ranks changes for them has already been disappointed, and being disappointed by it early is cheaper than being disappointed by it after the money leaves.

What is a product innovation, and what is a process innovation?

Product Innovation

A product innovation changes what the buyer receives. Something about the thing sold is different: what it does, what it is made of, how long it lasts, what it comes with, or what the buyer no longer needs to do for themselves. The definition is that short, and the test that goes with it is a question that could be put to the person paying.

Would the person paying notice without being told? If the honest answer is yes, the change is a change to what is sold. A register that opens flat on the desk instead of springing shut is noticed on the first morning by whoever is writing in it, and nobody has to explain the improvement to them. The test is doing its work there.

Then the consequence, and it is the one the rest of this reading order leans on. The thing improved is the thing being sold, so somebody has to want the improved version enough to buy it. A change to what is sold has to be recovered out of volume nobody has bought yet. How that recovery actually runs, and how it compares with the other kind, is covered separately under Product Innovation vs Process Innovation: Which One Pays.

Process Innovation

A process innovation changes how the same thing gets made or delivered. The buyer receives what they were already receiving, at the counter they were already using. Run the same test from the other side and the answer flips: the person paying would not notice, and not being noticed is the defining feature of this kind rather than a limitation of it.

The consequence flips too. A change to how something is made is recovered out of volume that already exists. Recovery out of existing volume is why such a change can pay from the first week, and why it can pay nothing at all. Both of the works lifts are changes of this kind. Neither touches anything a school buying registers would ever see. One is worth 1,00,000 registers of yearly ability and the other is worth nothing. The two kinds are not a ranking but two positions, and the return lives in the position rather than in the kind.

Now the sentence that stops anybody treating a change to how something is made as a solved problem. Fixing a constraintWhatever is holding the whole thing back, in the sense that moving it moves the result and moving anything else does not. Finding it is covered separately, step by step. relocates it rather than removing it. Cutting at 150 and printing at 125 can both keep up. Take binding to 125 an hour and the system rateThe rate a whole line of stages runs at, which is the smallest of the individual stage rates rather than their total or their average. It is worked out from the stage rates, covered separately. would go to 125. But printing runs at exactly 125 as well. The two would tie, and passing 125 an hour from there would need printing and binding raised together rather than either one on its own.

Widening the narrow doorway of a shop moves the queue to the till. Nothing was wasted; the doorway genuinely was the problem. A queue is simply a property of the whole arrangement, so it goes wherever the arrangement is now tightest.

What does a change in method do to the economics of a whole field?

How Innovation Changes Industry Economics

Three things move when a method changes, and they are worth taking one at a time rather than as a single blur about progress.

First, the ceiling. A works’s yearly ability to supply moves without anybody having bought a single extra order, and the four rows demonstrate exactly that. A field whose ceiling has moved can supply more than it could before, whether or not anybody wants more. The moved ceiling is a fact about supply, and it arrives before any fact about demand.

Second, the shape of the cost. A change in method usually converts cost that moved with volume into cost that stands still, and the standing kind has to be earned back out of volume every single year, in good years and thin ones alike. Anjani Stationers gives one published movement to anchor that on. Its standing base of Rs 49,60,000/- against revenue of Rs 2,40,00,000/- is 20.67 per cent, and the following year Rs 74,00,000/- against Rs 2,70,00,000/- is 27.41 per cent. The share of this business's revenue that stands still whatever happens rose 6.74 points in one year.

Now the honest thing, said straight away. Nothing published anywhere says what caused it. The two percentages are divisions of published figures, and the rise between them is arithmetic on those divisions. Where the money sits can be read off the face of the statement, in employee benefits, in other operating costs and in depreciation. The purpose of the money is the separate question, and no account answers it. Writing an answer in would supply the one fact nobody disclosed. The signature is visible in the accounts; the cause is not.

Third, who can afford to stand in the field at all. A higher standing base means a higher volume needed simply to hold position, and every business in a field where methods have shifted the same way carries the same arithmetic. The arithmetic is a statement about the shape of the field rather than about any one business’s judgement. A change in method is visible in the accounts as a change in what stands still, long before it is visible as anything else, and the accounts do not say what the change was.

What is an incremental change, and when does a run of small ones stop being small?

Incremental Innovation

An incremental innovation keeps the same idea and improves it: the same thing sold to the same buyers, made in broadly the same way, only better. A sharper blade, a tighter layout on the sheet, a jig that saves a movement. Two honest things need saying about it, and they need saying in the same breath rather than one at a time.

The first is that this is where almost all of the work is. A business improving what it already runs is doing the ordinary thing, and doing the ordinary thing well is not a lesser activity than doing an unusual thing badly. Most of what any works gets better at over a decade arrives this way and arrives without a name.

The second is that its return obeys exactly the same rule as everything else here. Incremental describes the step and not the return, and the two are routinely confused. An incremental change at the stage setting the rate would be worth 1,00,000 registers of yearly ability. An incremental change of the same size one stage over is worth nothing. Nobody would call the second one large, and nobody should call it a return either.

Which brings the accumulation question, and it deserves asking properly rather than being waved at. A run of small changes can add up to a large movement. A run of small changes can also add up to nothing at all. Which of the two it is depends on whether the changes kept landing at the stage that was setting the rate at the time, and since fixing that stage relocates the constraint, the target moves between one improvement and the next. A household putting a little aside each month compounds only if the money survives the month.

How is that different from the kind that unseats the people already there?

Incremental vs Disruptive Innovation

Size is the instrument that just failed twice, so the two kinds separate on the axis already built and not on size. An incremental change improves what is already there. An incremental change is judged by the same people against the same measures, it arrives in the same shape as the existing work, and it is therefore visible to everybody in the field at once. Somebody is always assigned to it.

A change of the other kind arrives by a different route. A change of the other kind is judged against measures the established business does not use. Such a change usually starts by serving somebody that business was content to leave alone. And it is therefore invisible on the instruments that business actually reads, not because anybody is careless but because no dial was ever built for it. The separation is not small against large, it is visible against invisible on the measures already being watched.

The account of how an established business loses to a change of the second kind was set out by Clayton M. Christensen in 1997. Why serving the most profitable customer better is the correct decision each year and the wrong decision across years is covered separately under Disruptive Innovation: Why Good Companies Lose.

The separation stands on its own, without a case attached to it.

Which change actually threatens a business that is already established?

An established business is threatened by the change it can see least, and not by the change that is largest. The sentence carries a great deal of work, so it is worth earning three separate ways rather than asserting it once.

Take it on the works first. The stage everybody watches is the one running fastest and looking busiest, and the stage setting the rate is the one standing still with a queue in front of it. A proposal ranked by visibility spends at printing, every time, and the arithmetic above says what that buys. The visible stage and the decisive stage are two different stages, and nothing on the shop floor announces which is which.

Take it on attention next. A large change arrives with a name, a trade press writing about it and a competitor everybody already tracks, so it is met. A change that does not resemble the work at all arrives with none of those three. Coming up at a meeting requires somebody whose job it is to raise it, and nobody in the building has that job.

Take it on measurement last. Measurement is the hardest of the three. A business measures what it decided to measure some years ago, when a different set of questions mattered. A change that scores badly on those measures reads as unimportant, and it goes on reading as unimportant right up until the measures turn out to be the wrong measures. The practical question is not which change is biggest, it is which change a business's own instruments would fail to register.

A shop watches the shop across the road, the one it can see from the door. The one that takes its customers is two lanes over and sells something slightly different to the same people, and nobody in the first shop has ever walked past it.

Can any of this be read off a set of published accounts?

How to Assess a Company’s Innovation Capability

A business’s capability at changing its methods cannot be assessed from published accounts. The refusal is the first line rather than the last. Softening it would waste the reader’s time, and the reader who arrives wanting a method is the reader most likely to build one out of whatever is lying around.

The refusal turns into a finding rather than a shrug once somebody writes down exactly what would have to be disclosed before the assessment could be made. Six things, each one a line, and each one a thing somebody inside the business knows perfectly well.

1. What was spent on changing the method, separated from what was spent on doing more of the same thing.
2. Which of the two kinds each spend belonged to: what is sold, or how it is made.
3. When each spend was made, and when the revenue it was meant to produce arrived.
4. What was attempted and abandoned, because a count of successes with no count of attempts is not a rate.
5. Who did the work, and whether they are still there.
6. What the business would have earned had it done nothing, which is the comparison every claim of capability quietly assumes and nobody states.

Now land it on the case. Anjani Stationers discloses none of the six. Its standing base rose Rs 24,40,000/- in a single year, a rise of 49.19 per cent, on people, on space and on a binding operation the business bought into. The three heads sit on the face of the statement and each carries a figure: Rs 6,00,000/- against employee benefits, Rs 11,40,000/- against the part of other operating costs that stands still, mostly a second warehouse, and Rs 7,00,000/- against depreciation and amortisation on the assets bought. The three of them close on the rise exactly. Nobody disclosed the split itself, and producing it took an estimate rather than a reading. Having the three figures still fills not one of the six lines from anything on the record, and no arithmetic on the published figures fills them either, because the figures carry totals rather than purposes.

And the plain thing about that matters for how the next hundred businesses get read. Disclosing none of the six is the ordinary situation and not a special failing. Most businesses disclose none of the six, the filing regime does not ask for them, and a reader who produces a capability judgement anyway has produced a judgement about their own confidence rather than about the business. A spend made now for revenue later changes a reported year, and that change is covered separately under Research and Development: Spending Today for Revenue Later.

Which part of any of this is local, and which part is not?

Almost none of it is local, and saying so directly is better than leaving a reader to wonder. India supplies the way money is written, the way digits are grouped, the legal form Private Limited carried by the invented business, the 250 working days sitting behind the line-hours, and the school year that decides when registers get ordered. Nothing else here belongs to any one country.

The mechanism is universal, and that is arithmetic rather than flattery. A line of stages makes what its slowest stage makes in every works on earth, and a business watches what its own instruments read in every field on earth. The mechanism turns on no rate, threshold, period or statutory definition, so no authority anywhere is the authority on it.