Technology Adoption and Diffusion: What Sets the Pace
Pace belongs to the people deciding, and never to the change itself. Each of them weighs whether it beats what is already on their own bench, whether it fits, how much has to be learned, whether it can be tried in a small way and whether a result would be visible to anybody else. Everett M. Rogers set that frame out in 1962. One buying relationship gives a duration, and a duration is not a pace.
Two businesses put out the same change on the same day. Why does one of them move?
Most people meet technology adoption first as a drawing. The shape that taking up something new makes over time, what sits along the bottom and up the side of it, why the line bends where it bends, the five positions strung along it in their order of arrival and the full list of what anybody would have to hold before drawing one honestly are all covered separately under The Adoption Curve: How New Products Spread.
A different question is left over. Suppose the change exists, suppose somebody could take it up, and suppose nobody has any drawing to consult. What decides how quickly the taking up happens, or whether it happens at all? The question is one of pace.
The awkward pair belongs on the table straight away. Two businesses put out a change of the same kind on the same morning. One of them is taken up within the season. The other is left exactly where it stood. The reasonable instinct is that the first change must have been the better change. Nothing separating those two outcomes lives inside the changes themselves.
A version of the pair can be stood in front of. Two lanes of the same market, selling the same goods, to the same people walking between them. In one lane every stall took to a new way of accepting payment inside a single season. In the other lane not one stall did. Same goods, same customers, same season, and an answer sought in the machines would be sought in the wrong place entirely.
Two businesses put out the same kind of change on the same day. One is taken up quickly and the other is not. Where should the search for the difference begin?
Diffusion of Innovation: what does the word actually describe?
Diffusion is the spread of something new through a set of people who could take it up, watched over time. The plain meaning stops there, and the word still does a lot of quiet work. Diffusion turns a great many separate decisions into one noun, and a noun sounds like a thing that happens on its own.
Two things have to be in hand before anything quantitative can be said about that spreading, and both are requirements rather than caveats. The first is a defined set of people who could have taken the thing up: named, bounded, counted once each. The second is a date for each one who did. Without both of those, there is a subject and there is no measurement. Description, argument, comparison and reasoning can go on all day. A pace cannot be stated.
The frame around all of this belongs to somebody. Everett M. Rogers set it out in Diffusion of Innovations in 1962. He described what an adopterThe person or the business doing the deciding. Somebody with arrangements already in place that a new thing would have to fit into, rather than a category anybody belongs to permanently. weighs and gave the spreading its vocabulary. He did not hand anybody a rate, and a book is not a source for a number.
What two things are needed before anybody can state how fast something is spreading?
So what does the person deciding actually weigh?
Take them one at a time, and phrase each as the question the person is silently asking rather than as a heading in a book. None of the five carries a weight, a score or a place in any order of importance.
The first is relative advantage. Is this better than the thing already sitting on my own bench? Notice where the comparison is being made. The seller compared the change against whatever the seller chose, most likely the common version across the whole tradeA line of business together with the people who work in it. Used here for the group somebody swaps notes with, rather than for any single act of buying or selling.. The person deciding compares it against one particular thing in one particular place. Relative advantage is measured against what the adopter already runs. A change that is plainly better in general can be worth nothing in particular.
The second is compatibility. Does it fit what I already do, what I already hold and what I already believe about how this work goes? A change can be better on every measure anybody publishes and still sit awkwardly beside three arrangements that were built around the old way of doing it.
The third is complexity. How much do I have to learn, and how much of that learning falls on people who never asked for any of this? That second half is the part sellers rarely count. The learning is real work and somebody has to do it while the ordinary work carries on.
The fourth is trialability. Can I try it in a small way, and can I stop? The fifth is observability. If it works here, will anybody outside be able to see that it worked? Hold on to those two.
The frame itself is not an instrument. The five are a list of what gets weighed rather than a set of weights. Nothing in the frame says which one matters most, nothing scores anything, and a version of this list with numbers printed beside it would be a measurement nobody ever made, wearing a borrowed name.
Stand it in a kitchen. A household is offered a faster way of heating water. The question at the table is not really whether it is faster. The question is whether it fits the shelf it has to live on, whether the person who does the cooking has to relearn anything, and whether the thing can be sent back if it turns out to be a nuisance.
Now watch the first of those five do something that looks like a contradiction and is not. Draw the identical change twice, and put it against a different bench each time. The change has not altered by so much as a millimetre between the two pictures. The verdict has reversed completely.
A change is measurably better than the version most of the trade uses. One business looks at it and decides it offers nothing. Under the frame set out here, what has most likely happened?
Why do the two the seller controls least matter most?
Trialability and observability behave as a pair, so take them together. Sellers usually treat both as things to mention in passing. The people deciding treat both as the decision itself.
A change that can be tried in a small way converts a judgement into an experiment, and that conversion is worth more than it sounds. An experiment carries a way back out of it and a commitment does not, so a business will run an experiment it would never authorise as a commitment. Take the small first step away and the verdict is being asked of somebody who has no way of finding out.
A change whose result can be seen converts one adopter into evidence for the next. Take that away and every single person has to reach the same conclusion alone, from nothing, with no help from anybody who went before. A change that cannot be tried in a small way is asking for a decision, and a change that cannot be seen working is asking for the same decision again from every single person.
The same rule appears again with a person standing where a stage stood. Under throughput, Anjani Stationers Private Limited, an invented maker of stationery, runs a worksThe building and the machinery where making happens, as against the office where the paperwork is done or the shop where the selling is done. that cuts, prints and binds, and the throughput working shows the works making 100 registers an hour with a rated capacityThe most a place could turn out in a year with everything running as intended. A rated capacity is a ceiling on ability rather than a record of what was actually made. of 4,00,000 registers a year. Lifting binding, the slowest of the three stages, would buy 1,00,000 registers of yearly ability. Printing already runs faster than binding, so lifting printing would buy exactly nothing. Only the position decides. On the buyer's side, the position that decides the return is the adopter's own, and nobody outside can read it off a specificationThe written statement of what a thing does and how well it performs. A specification describes the object completely and says nothing at all about the place the object would have to fit into..
A shopkeeper makes this concrete in about four seconds. He will take a new till for a week on trial. He will not sign for one. And the reason he asked for the week in the first place is that the shop two doors along visibly stopped queuing.
Which pair of the five is singled out as the pair a seller controls least and reports on never?
What carries a change from one person deciding to the next?
A change reaches the next person through something, and the something is usually other people who have already decided rather than the seller. Two things do most of that carrying, and neither of them is a property of the change.
The first is who says it. Somebody running the same work, with nothing riding on the answer, is worth more than any account the seller can give. Credibility is doing the work there, not persuasion. The person listening has no way of separating a truthful seller from a hopeful one, so the seller may be entirely truthful and still be discounted.
The second is how closely connected the people who could take it up already are. A trade where everybody meets at the same market, compares notes at the same counter and can walk past each other's premisesThe buildings a business works out of. A passer by can see how the work is actually being done at a premises. A written account cannot show that. behaves quite differently from one where nobody ever does. The pace of a change is partly a fact about the people it is travelling among. The same change moves at two speeds in two trades.
Ten shops sharing one small mall see each other's counters every working day. Ten shops spread across a district never do. Putting the same change in front of both sets changes nothing about the change and a great deal about how far each result can travel.
Nobody has ever measured what either carrier is worth, so the carriers can be named and no rate, no lag and no proportion attaches to any of them.
Two trades are handed the identical change. In one, everybody meets at the same market every week. In the other, nobody ever meets anybody. What does that difference belong to?
Four shapes look identical when drawn. What does each one count?
Four rising S shapes appear across these notes. All four climb slowly, then quickly, then slowly. All four are usually drawn with nothing written on either axis. Set side by side with the labels removed, nobody alive could sort them.
One counts sellers in a field over time, and how a field's own arrival, crowding and thinning are read is covered separately under The Industry Life Cycle: Emergence to Decline. One counts buyers who have taken something up, added together as it goes, and that is the shape the subject of this guide belongs to, covered separately under The Adoption Curve: How New Products Spread. One counts a single product's own volume and margin from launch to withdrawal, covered separately under The Product Life Cycle: Launch to Withdrawal. One counts performance against effort spent, covered separately under The Technology S-Curve: Why Progress Slows and Then Jumps.
Four shapes that look identical drawn are about four different things, and nothing about the drawings themselves reveals which one is in hand. What is being counted along the bottom and up the side has to be read before anything else in the passage the drawing sits in. If the drawing does not say, it has taught nothing yet, and the person who put it there may not have decided either.
Four shapes in these notes look identical when drawn. Which one counts buyers who have taken something up?
So what is actually held about anybody taking anything up?
One relationship, one duration, and it deserves to be worked honestly rather than waved at. The Sunrise Public School group has bought from Anjani Stationers Private Limited for eleven years and stands at Rs 81,00,000/- in the published year, which is 30.00 per cent of that year's revenue of Rs 2,70,00,000/-, sitting inside a book of 36 accounts.
The relationship supports something real, so start there. The relationship lasted eleven years. The eleven-year durationA length of time. A duration answers how long something has been running, and never how many of anything there are. is what makes the age of the arrangement readable at all, and somebody who knows the trade can weigh eleven years against what they know of it.
Now say what it does not support, item by item. The relationship does not say when the school group first took anything up relative to anybody else. It does not say who else took the same thing up, or when they did it, or how many of them there were. Nor does it say which set of buyers could have. One buyer, one date, and one point is not a path.
The remainder is worth having. The honest output is the duration, written as a duration, with the count of relationships beside it. Eleven years and one buying relationship can be handed to somebody who knows the trade and argued about properly. There is nothing inside a stated pace to disagree with, so it cannot be argued about at all.
Every control this subject invites would either put a value on a shape that nothing supports, or attach a weight to a frame that supplies none. The one honest control moves the number of dates held and shows how little agreement survives, and it sits under The Adoption Curve: How New Products Spread.
One buyer has bought for eleven years and is 30.00 per cent of the year's revenue. What does that support?
What would a pace claim need, and what is held against each requirement?
A gap that can be seen is worth more than a gap somebody describes. Five things would have to be in hand before anybody could state how fast something is being taken up, and only one of the five has even a partial answer.
| What a stated pace requires | What these notes carry | Why |
|---|---|---|
| A defined set of people who could have taken it up | nothing | Nothing anywhere defines one, and nobody publishes the boundary of such a set |
| A date for each one who did | one date, for one buyer | The Sunrise Public School group's eleven years is the only duration attached to anybody taking anything up |
| A count of those who have | nothing | Counting requires a list, and no list exists |
| A share of that set | nothing | A share needs a bottom half to divide by, and the first row is empty |
| A rate of change | nothing | A rate needs a run of counts across time, and there is one observation |
Five requirements and one partial answer between them. No pace can be stated in any form. Notice that the second row is partial rather than met: the requirement is a date for each one who did, and one date is not a date for each. The denominatorThe bottom half of a fraction: the whole that a part is being measured against. Anything expressed as a share needs one, and it has to be somebody's stated boundary rather than a round number. that the fourth row needs is exactly the thing the first row says nobody has.
The empty rows are a finding rather than an apology. Empty rows are the ordinary situation for almost any change anybody will ever be asked about. The set of people who could take a thing up is rarely defined by anybody. The dates are rarely held. A pace stated without them was produced by somebody who filled both gaps quietly, and the filling is invisible by the time the figure arrives.
A claim arrives that a change is being taken up quickly, with no set of possible adopters named and one date behind it. What is the honest output?
The change that was called slow to catch on, when nobody had asked anything of the people it was meant to catch on with
A business puts out a change. Take-up is thinner than anybody expected. The review meets, and the question written at the top of the paper is why the change did not land. Four answers come back, and every one of them is about the change: it needs to be faster, it needs another feature, the price is wrong, the description is unclear. A second round of work is authorised, aimed squarely at making the thing better.
Nobody in that room asked any of the five questions the people deciding were actually weighing. Nobody asked what each of them was comparing it against. The comparison is whatever sits on their own bench rather than whatever the seller chose. Nobody asked whether it fits alongside what those people already do. Nobody asked how much has to be learned and by whom. Nobody asked whether it could be taken for a week and given back. And nobody asked whether, if it worked, anybody outside would be able to see that it had.
The tempting diagnosis is wrong, so say exactly what happened. Nobody was lazy and nobody misread a number. The room asked the only question it had evidence for, and that question was about the thing it built. The pace lives entirely outside the building.
The cost lands somewhere specific. The second round improves the one attribute that was never the constraint, so it spends another year and produces the same take-up. The result now looks confirmed rather than misdiagnosed, and that makes the third year harder to argue for than the second was.
The same error appears under throughput, worked inside a works with a person standing where a stage stood. Spending at the stage that was not setting the rate, and improving the attribute that was not being weighed, are one error in two settings, and neither can be caught by looking harder at the thing being improved.
The fix is one line and it is not a better product. Before anything is improved, the record should show what the people whose take-up was wanted were comparing it against, and whether they could have tried it without committing to it.
What should be asked of any claim that something is spreading quickly?
Four questions, put in this order, and the second one settles most of them
An analyst reading a strategy paper, a lender reading a business case or somebody inside a business reading their own review has about a minute, and four questions fit inside it.
First, among whom is it spreading? Push for a described set of people rather than the name of a market. A market's name is a label; a described set is a boundary somebody drew and can defend. If the reply is the name of a trade, the first requirement is still missing.
Second, how many dates are actually in hand? This is a count and the reply has to be a number. Somebody who says the evidence is extensive has not replied. Somebody who says four has, and four may well be enough to start arguing about.
Third, what are the people deciding comparing it against? Not what the seller compared it against. The comparison that actually gets made is against whatever sits on each of their own benches, and nobody writes that one down.
Fourth, could they have tried it in a small way, and could anybody see the result? This is the pair a seller controls least and reports on never, so it will almost always have to be asked for rather than read.
A claim about pace with the first two answers blank is an impression rather than a finding. The second question does a great deal of work on its own. Exactly one date is held, and a count of one can be established in about four seconds. The second question alone stops the claim.
Of the four questions asked in order, which one would on its own stop a pace claim?
Does any of this belong to one country?
Almost none of it. India supplies the way money is written, the way digits are grouped, the legal form carried by one invented business and the school year behind the one published relationship. What a person weighs before taking something up is the same everywhere, and so is the requirement for a described set of people and a date for each one who acted. No rate, threshold, period or statutory definition appears here for any authority to be the authority on.
Where the one borrowed name gets checked
| Source | Document | How it is treated here | Where |
|---|---|---|---|
| Everett M. Rogers | Diffusion of Innovations, 1962 | Cited for the frame that names what an adopter weighs, and for the vocabulary of something spreading through a set of people who could take it up. | worldcat.org |
| Fin Maverick teaching notes | The built working quoted above, on throughput and on how concentrated a book of buyers is | The only figures printed above are one duration, one share of one year's revenue and one count of buying relationships, each carried across from earlier working rather than reworked. | finmaverick.com |
Anjani Stationers Private Limited and the Sunrise Public School group are invented.
Educational material. Not advice on any investment, tax, budget or market position.