The lightsaber existed for about forty years before the laser did. The glowing blade shows up in pulp science fiction in the 1930s. It becomes the Jedi's weapon by 1977. It lodges in millions of imaginations, decades before the physics could have built anything like it.
The belief ran ahead of the thing. And the belief was not idle. The man who built the first working laser said he was set on the path by reading a death-ray novel as a boy. The fiction did not describe the invention. It recruited the inventor.
That is the order of operations most people in revenue get backwards. We are trained to think the product comes first and demand answers it. You build the thing, then go find the people who want it.
More often than is comfortable, it runs the other way. The belief that a thing should exist, that a problem matters, that it must be solved now and solved this way: that belief is the demand. It is manufactured, not discovered. The engineering chases the contagion.
So if you sell for a living, the question is not how to fulfil demand. It is how to start the contagion.
The art and the order-taking
Split the market in two. There is demand generation: moving a group of people from not believing they have a problem to believing they have your problem. And there is demand fulfilment: taking an order from people who already believe.
Fulfilment is necessary. Generation is the art. The whole profession has drifted toward fulfilment, because fulfilment is legible. You can put it on a scorecard. You can break it into stages, score the stages, forecast the total.
I have argued in Self-Report Is Not a Measurement that we do even this badly. We ask the seller what they believe instead of reading what the buyer did. But the deeper problem is that the scorecard only measures the order-taking.
It starts after the belief is already set. It says nothing about the one thing that decides most deals before a rep is in the room: who shaped what the buyer believes is true.
Enterprise selling, done as an art rather than a clerical function, is paradigm intervention in a community of practice: the deliberate work of shifting what a connected group of people holds to be true, in your favour.
That is not a metaphor for selling. It is a definition of it. Everything else is admin.
How a belief travels
If demand is a belief spreading through a group, then generating it is a question about how beliefs move through groups. That has been studied, exhaustively, by people who never sold anything.
The first finding: a group is not a set of individuals deciding privately. It is a single thing that stores and transmits belief collectively. Joseph Henrich calls it the collective brain. Its capacity to hold and grow an idea scales with how many minds are connected, and how well.
The proof is in the reverse. When Tasmania was cut off from the mainland by rising seas, its population did not just stop gaining ideas. It lost them. The toolkit regressed to something simpler, because there were no longer enough connected minds to hold the complex one.
Disconnect a community and belief decays. That is not a fact about prehistory. It is the mechanism running inside every account, every industry, every buying group you have worked.
The second finding is how belief crosses between people, and it embarrasses the instinct to flood the market. Your close, dense contacts mostly already believe what you believe. A tight cluster recirculates the same conviction.
New belief arrives through the bridges into another cluster. What makes a tie a bridge is the structural hole it spans, not whether the tie is weak or strong. Granovetter saw the pattern first in weak ties. Ronald Burt located the mechanism in the hole itself, and noted that strong-tie bridges are common too.
The people who sit across those holes, the brokers, get first sight of the new idea and disproportionate credit for carrying it. A new idea enters a community not through the front door of the dense core, but through the broker at the edge who trades in novelty for a living.
The third finding tells you where the next belief will catch. It concentrates in a hothouse, then spreads through a diaspora. The University of Utah's computer-graphics lab in the 1960s scattered its people into Pixar, Adobe, Silicon Graphics and the modern GPU.
Shockley's defectors built Fairchild, and Fairchild seeded most of Silicon Valley. The eight authors of one 2017 paper on machine attention have already scattered into a dozen companies carrying the same idea.
The pattern is scale-invariant. It runs the same at the level of a civilisation and at the level of a sales methodology. MEDDIC is a hothouse-and-diaspora story: codified inside one company in the 1990s, then carried outward by the thousands of sellers who trained there, until it became the lingua franca of enterprise sales. Nobody mandated it. It spread like the contagion it was.
The membrane is language
Here is the operational hinge, and it is the part most sellers never touch. A community of practice, and a buying group is one, is bounded by its language. Its jargon, its exemplars, its war stories are the membrane.
Speak the dialect and you are admitted to the room where belief forms. Fumble it and you are marked as an outsider whose ideas can be ignored.
Thomas Kuhn made the point about science, but it holds for any group with a shared way of framing its problems. The accepted vocabulary is the paradigm. Whoever supplies the words supplies the conclusions you can reach with them.
A security team that frames its problem as "audit-scope reduction" reaches different decisions from one that frames it as "data-centric defence". Same product, same facts, different governing sentence, different outcome.
The unit you are selling is not a product and not a feature. You are selling a sentence: a reframe of the problem that, once a community adopts it, makes your solution the obvious answer.
This is why being right is not enough, and never has been. Mendel was right in 1866 and ignored for thirty-four years, because the community had no language to receive him. Darwin took credit Wallace deserved a share of, partly because he was better connected.
Recognition is a property of the network, not a clean readout of merit. The corollary for revenue is bracing. The better product loses, routinely, to the one whose makers supplied the community's vocabulary first. Demand generation is the contest to supply that vocabulary before anyone else does.
Reading the room you are not in
The uncomfortable modern fact is that you are mostly not in the room. By Gartner's reckoning a buying group spends only a sliver of its time with any vendor. Split across all of them, that leaves each seller a single-digit share of the buyer's attention. Most of the deciding happens in conversations you will never attend.
You cannot be present for the formation of the belief. So the work moves upstream. You have to pre-load the community's vocabulary, so the frame it reaches in your absence is already yours.
To do that you have to read the paradigm a buyer thinks in before you meet them. And that is readable, because people inherit their frames from the communities they have passed through. Every employer, every training lineage, every certification, every regulatory regime leaves a residue of vocabulary and belief.
Trace those memberships and you have a prior on how someone will frame the decision before they say a word. I have written elsewhere about building this as an instrument: a co-presence graph over career histories that infers which communities of practice a person belongs to, and how strongly. The same machine that finds the contact reads the dialect they speak.
One caveat the word contagion earns, before you lean on any of this. When a belief spreads across a community you can never fully separate influence from likeness. People who already resemble one another adopt the same frame without catching it from anyone (Shalizi and Thomas, 2011). So contagion is the working model and a predictive prior here, not a proven mechanism. It earns its keep by telling you where to look, not by certifying cause.
A few other things have to be right, or the instrument lies to you. The weighting is not simple recency. The first paradigm a person was trained in, and the most intense and prestigious rooms they have sat in, print far deeper than last week's webinar. Model the imprint, not the timeline.
The path through the community matters more than the headcount. Enter through the broker at the edge who welcomes a new frame. Win the endorsement of the embedded guardian who confers legitimacy. Then let the dense interior copy them.
And the model is a prior, never a verdict. The moment the buyer's own words contradict the profile, the words win. A provenance model that overrides what the person in front of you is telling you is not intelligence. It is prejudice with a graph attached.
The reframe itself has a design constraint that catches most people out. Mirror a community's language perfectly and you are admitted, but you have said nothing new. Flawless fluency is, by construction, the absence of novelty. And novelty is the only thing that moves a belief.
So you need both halves. Enough of the dialect to be trusted, and a reframe transmissible enough that your champion can repeat it, in their own words, in the room you are not in. A reframe only you can articulate dies in the meeting. A reframe your buyer can carry, and claim as their own insight, propagates. Let them have the credit. Their credit is your distribution.
What I learned selling Protegrity
This is not theory I assembled after the fact. When I took Protegrity across APAC, the job was never to fulfil demand for data protection. It was to change what a security community believed the right protection was, and which data even warranted it.
That community was not one room. It was a federation. A compliance-led group that framed everything as audit scope and defensibility. An architecture-led group that framed it as threat coverage and defence-in-depth. An operations-led group that framed it as latency and blast radius.
The same underlying reframe, protect the data itself, at the field, without a vault, had to be re-lexicalised three times, once into each dialect. It had to be entered through the architects who bridged the others, and ratified by each group's guardian before it would spread.
When it worked, it did not feel like closing deals. It felt like a belief catching across a market, and the orders followed it the way they always do. The contagion was the work. The signatures were the by-product.
The idea eats itself last
There is a final turn, and it is the one that tells you the model is true. This essay describes how beliefs spread through communities. If it spreads, it will spread by exactly the mechanism it describes.
It will radiate from somewhere small, carried by people with weak ties to the sales world and a foot in the data world, who can broker between the two. It will travel as language. Its origin will be argued over.
And if it ever wins, its vocabulary will go generic. Unownable, common property, the way MEDDPICC did: ruled by a court this year to belong to no one, because by then it belonged to everyone. That is not the failure of an idea. It is the last stage of its success.
You do not need more demand to fulfil. You need to stop mistaking fulfilment for the job. The pipeline you are fighting over was created upstream, by whoever shaped what your buyers believe is true. That work is available to you, earlier than you think, if you treat the belief as the product and the order as the receipt.
The operating question is not "how do we close this deal?" That asks about the order. It is: whose sentence is the buyer using to frame the problem, and is it ours, spreading in the rooms we will never enter?
This is the upstream companion to Self-Report Is Not a Measurement, which reads the buyer once demand exists, where this one manufactures the demand. It runs on the instrument described in the co-presence-graph working paper, and shares the collective-brain and weak-tie logic with the flagship, The Deal Was Perfectly Qualified. It Still Died.