What Crossing the Chasm Actually Says
In 1991, a consultant named Geoffrey A. Moore published a marketing book aimed at a narrow crowd: people trying to sell high-tech products that were too new for most buyers to understand. The book was Crossing the Chasm, and it went on to sell more than a million copies, with revised editions in 1999 and 2014. Three decades later, founders still argue about it in board meetings.
Moore's core claim is simple and a little brutal. The way a breakthrough product wins its first fans is the exact opposite of the way it wins the mass market. The early crowd and the mainstream crowd are not the same people at different times. They're different people with different psychology, and the transition between them is not a gentle slope. It's a gap wide enough to swallow companies whole.
That gap is the chasm. On one side sit the technology enthusiasts and visionaries who chase anything new. On the other sits the pragmatic majority who wait until a product is boring and proven. Plenty of startups get a burst of early traction, mistake it for a real market, scale up their spending, and then watch growth flatten. They never fell short on the product. They fell into the chasm.
Moore built his idea on top of an older one. In 1962, sociologist Everett Rogers published Diffusion of Innovations, which described how new ideas spread through a population in a predictable bell curve of adopter types. Moore's contribution was to argue that the curve lies. It looks continuous, but for disruptive tech products there's a hidden crack in it, and knowing where the crack is changes everything about how you go to market.
The Technology Adoption Life Cycle
Rogers' model splits any market into five groups based on how quickly they'll adopt something new. The percentages come from the statistical shape of a bell curve, and they're worth memorizing because they frame the whole problem.
| Segment | Share of market | What they want | Buying trigger |
|---|---|---|---|
| Innovators | 2.5% | The technology itself | It's new and technically interesting |
| Early adopters (visionaries) | 13.5% | A competitive breakthrough | A bold vision they can champion |
| Early majority (pragmatists) | 34% | A safe, proven productivity gain | References from peers they trust |
| Late majority (conservatives) | 34% | A standard everyone else uses | It's now the established default |
| Laggards | 16% | To be left alone | Forced to, or never |
Read left to right and you can see the story. The first two groups, innovators and early adopters, make up only 16% of the market combined. That's the early market. The middle two groups, the early and late majority, are 68%, which is where the money and the durable business live. That's the mainstream market.
Moore's insight is that the boundary between the early adopters and the early majority is not a normal step up the curve. Every transition has a small crack in it, a moment where one group's reasons for buying don't quite carry to the next. But the crack between visionaries and pragmatists is a canyon. The reasons a visionary buys have almost no pull on a pragmatist, so word of mouth stops working right when you need it most.
That's why Moore drew the curve with an actual gap cut into it. The picture matters because it corrects a dangerous instinct: the belief that if you just keep doing what got you your first customers, the rest of the market will follow. It won't. The early market and the mainstream market are separated, and you have to cross on purpose. The same "listen to your traction and scale it" instinct is what dooms many great companies, a pattern we cover in The Innovator's Dilemma.
Why the Chasm Exists: Visionaries vs Pragmatists
To cross the chasm, you have to understand why it's there, and that comes down to two buyer psychologies that barely overlap.
Early adopters are visionaries. They don't buy a product; they buy a future. A visionary sees a new technology and immediately imagines a strategic leap, a way to leapfrog competitors, a project that could make their career. They'll tolerate bugs, missing features, and thin support, because they expect to do a lot of the work themselves. They want to be first. Being first is the whole point.
The early majority are pragmatists, and they want almost the opposite. A pragmatist doesn't want to be first; they want to be right. They've watched too many shiny technologies turn into expensive dead ends, so they wait. They buy to solve a concrete problem with a proven tool, not to chase a vision. And here's the trap: the single most important thing a pragmatist wants is a reference from another pragmatist in the same industry, someone who already bought the thing and is glad they did.
| Dimension | Visionaries (early adopters) | Pragmatists (early majority) |
|---|---|---|
| Motivation | Get ahead, make a leap | Keep up, avoid falling behind |
| Attitude to risk | Risk is opportunity | Risk is to be eliminated |
| Product expectation | A promising core they can build on | A complete, supported solution |
| Who they trust | Analysts, technologists, their own vision | Other pragmatists in their own industry |
| Buying pace | Fast, on conviction | Slow, on evidence and references |
Now the chasm makes sense. To sell to a pragmatist, you need references from other pragmatists. But before you cross, all your happy customers are visionaries, and a pragmatist does not consider a visionary a credible reference. In their eyes, visionaries are the reckless people who buy things that don't pan out. So your existing customer base, the one that got you this far, is worthless as social proof to the people you need next. You're standing on one side of the canyon holding testimonials that the other side won't read.
The D-Day Strategy: Pick One Beachhead
Moore's answer to the chasm is the part founders resist most, because it feels like shrinking your ambition at the exact moment you want to grow it. His advice: stop selling to everyone and attack one narrow market segment with everything you have.
He calls it the D-Day strategy. On June 6, 1944, the Allies didn't invade Europe along its entire coastline. They concentrated overwhelming force on the beaches of Normandy, established an unbreakable foothold, and only then pushed inland. The goal wasn't Normandy. The goal was Europe. But you can't take Europe without first owning one beach.
For a startup, the beach is a single, tightly defined segment of pragmatist customers who share a pressing problem, talk to each other, and buy the same way. Moore's rule is to pick a target so specific that you can realistically become the dominant, obvious, must-have solution for that one group, and to become it fast. Not a leading option. The option.
This works because of how pragmatists buy. They trust references from people like them, and "people like them" means their own tightly connected niche. Win 6 of the 10 companies that matter in one small segment and word travels fast inside that segment, because those buyers all go to the same conferences, read the same newsletters, and call each other. Win 6 out of 600 companies spread across a dozen industries and you've made no dent in anyone's reference network. You've been a rounding error everywhere instead of a leader somewhere.
The math of focus is counterintuitive but ironclad. A dominant 70% share of a tiny market is worth far more than a 2% sliver of a huge one, because share inside a self-referencing niche compounds into word of mouth, and word of mouth is the only engine that carries you across. Founders hate this advice because it means saying no to real revenue from customers outside the beachhead. Moore's point is that scattered revenue doesn't build the reference base you need, so it's often a distraction dressed up as growth. This is the same "do the unscalable thing first" logic Paul Graham gives early founders, which we unpack in Do Things That Don't Scale.
The Whole Product Gap
Picking a beachhead is half the crossing. The other half is giving those pragmatists something they'll actually buy, and that's usually more than the product you think you're selling.
Moore separates the "generic product," the thing you ship, from the "whole product," everything a customer needs to get their problem fully solved. A visionary is happy to buy the generic product and build the rest themselves, because tinkering is part of the appeal. A pragmatist wants the whole thing to already exist: the core software plus the integrations, the onboarding, the documentation, the support line, the compatible hardware, the consultants who know how to install it, the community they can ask for help.
Think of it in layers. At the center is your core product. Around it sits everything required to deliver the full promise: setup, training, related tools, standards, partners, service. Pragmatists evaluate the whole thing. If any layer is missing, they read the gap as risk, and risk is the one thing they refuse to buy.
| Generic product | Whole product | |
|---|---|---|
| What it is | The core feature you built | Everything needed to fully solve the problem |
| Includes | The software or device itself | Integrations, support, training, partners, docs |
| Who accepts it | Visionaries who'll fill the gaps | Pragmatists who want zero gaps |
| Missing pieces read as | An interesting project | Unacceptable risk |
This is why crossing the chasm is rarely a solo act. You often can't build every layer yourself in time, so Moore stresses tactical alliances: partnerships that fill the whole-product gaps for your specific beachhead. The narrow segment helps here too. Defining the whole product for one type of customer is achievable. Defining it for everyone at once is not. Focus doesn't just win you references; it makes the whole product buildable in the first place.
How Documentum Crossed Its Chasm
The cleanest example Moore tells is Documentum, a document-management company that was stuck and nearly gave up on the idea of a broad market. Instead of chasing every company that had documents to manage, which is essentially all of them, they picked one absurdly specific beachhead: the regulatory affairs departments of large pharmaceutical companies.
The choice looked tiny. By Moore's account there were only about 40 such departments worth targeting, staffed by roughly a thousand people total. But those departments shared an intense, expensive, shared pain. Getting a new drug approved means filing a regulatory submission that can run to hundreds of thousands of pages, and every day of delay in approval is a day of lost patent-protected sales, often worth enormous sums. For that buyer, better document management wasn't a nice-to-have. It was money.
Documentum went all in on that one niche and dominated it. The revenue curve tells the story: the company grew from around $2 million to $8 million to $25 million to $45 million over roughly three years as it took over pharma. Competitors shrugged. The line on Documentum was, "Sure, they're good for pharmaceuticals, but that's it."
Then the bowling started. Pharmaceuticals are a process industry with heavy documentation and regulation, and it turns out other industries look similar from that angle. Oil and gas companies manage mountains of documents for leases and exploration. Financial firms document complex derivatives. Each adjacent market shared enough with the last that Documentum's references, whole product, and reputation carried over. The niche that looked like a ceiling was actually a doorway. Owning one beach let them take the coast.
The Bowling Alley and the Tornado
Documentum's expansion follows the two stages Moore says come after the crossing, which he later developed in his 1995 sequel Inside the Tornado: the bowling alley and the tornado.
The bowling alley is the phase right after you've won your beachhead. You knock down your first pin, the initial niche, and use its momentum to knock down adjacent pins: neighboring segments that share either the same customer type or the same whole product. A win in pharma document management sets up a win in oil-and-gas document management, which sets up the next. Each niche is deliberate and reference-driven. You're still focused, just widening the focus one adjacent pin at a time.
The tornado is what happens when the broader pragmatist market finally tips. At some point the early majority stops evaluating and starts stampeding, because the new thing has become the obvious standard and the real risk flips: now the danger is being the company that didn't adopt it. Demand explodes. The playbook inverts completely. In the bowling alley you win by being the perfect fit for a niche; in the tornado you win by shipping fast, grabbing share, and becoming the default before rivals do. This is the moment blitzscaling is built for, which we cover in Blitzscaling.
The danger Moore warns about is running the wrong playbook for your stage. Charge like it's a tornado while you're still trying to cross, and you'll burn cash spraying thin marketing across a market that isn't ready to stampede. Stay in cautious beachhead mode when the tornado has already hit, and a faster competitor becomes the standard while you're perfecting your niche. Knowing which phase you're in is a strategic decision in itself, closely tied to whether you've truly reached product-market fit.
Why AI Products Are Stuck in the Chasm
Moore wrote about a pre-internet world, but the chasm has rarely been more visible than it is with AI in 2026. It's the sharpest live example of the model in decades.
Consumer AI crossed almost overnight. Chatbots and image tools raced through the innovators and early adopters straight into mainstream use, because the whole product was trivial. You open a browser and type. No integration, no procurement, no training. When the whole product is that simple, the chasm narrows to a crack.
Enterprise AI is a different story, and the data is stark. A widely cited 2025 study from MIT found that roughly 95% of corporate generative-AI pilots delivered no measurable return. Surveys from firms like McKinsey and Menlo Ventures show the same split: nearly every large company is experimenting with AI, but only a small fraction, often cited in the single digits to low double digits, has scaled it into real profit impact. Adoption is everywhere. Crossing is rare.
That's a textbook chasm, and Moore's framework explains exactly why. The early buyers of enterprise AI were visionaries chasing a strategic leap, happy to run a rough pilot and fill in the gaps themselves. The mainstream buyer is a pragmatist who wants the whole product: reliability, security review, governance, integration with existing systems, audit trails, staff who know how to run it, and a reference from a peer in their own industry who already made it work. Most AI vendors shipped a powerful generic product and skipped every one of those layers. The model is brilliant; the crossing apparatus doesn't exist yet.
The way across is the same as it was for Documentum. Stop selling "AI" to everyone and pick a beachhead: one industry, one workflow, one painful and expensive problem where you can deliver a complete, trustworthy solution and rack up references pragmatists believe. The winners of this cycle won't be whoever has the most capable model. They'll be whoever crosses the chasm first, a shift we explore in The SaaSpocalypse.
How to Spot and Cross Your Own Chasm
The chasm isn't only a founder's problem. Anyone building anything, a product, a career, a body of work, a piece of writing, faces the same gap between early enthusiasts and a skeptical mainstream. The skill that matters is reading adoption honestly instead of getting high on early praise.
Early traction is seductive and misleading. A handful of passionate users, a viral demo, a wave of signups from people who love new things: none of that tells you the mainstream is coming. To tell real momentum from a visionary sugar high, you have to study how adoption actually plays out, across industries and across decades, and separate the signals that predict a crossing from the ones that flatter you. That's a reading and pattern-recognition habit, not a dashboard.
Here's where a deliberate learning system earns its keep. When you read case studies, essays, earnings calls, and post-mortems with Glasp's web highlighter, you're building a personal library of how real products crossed or fell in. Highlighting forces you to mark the actual mechanism, the beachhead they chose, the whole product they assembled, the reference base they built, instead of the tidy narrative written afterward. Over time those highlights become a searchable map of adoption patterns you can pull up the moment you face your own crossing.
The same discipline works for video. Founder talks, market breakdowns, and interviews with people like Moore hold most of the useful detail in offhand remarks, and they scroll past fast. Running them through YouTube Summary pulls out the timestamps and key points so you can capture the specific move a company made rather than a vague impression. And once your highlights pile up, Glasp's AI chat lets you interrogate your own notes, asking what beachhead strategies actually worked or which whole-product gaps sank comparable products, so your reading compounds into judgment.
Then there's the social layer, which maps neatly onto Moore's own thesis. Pragmatists cross when they see credible peers who've already crossed. Learning works the same way. Following how sharp thinkers in your field highlight and annotate through Glasp's community shows you which sources they trust and what they pull from them, giving you the same peer-reference advantage in learning that a good beachhead gives a startup. You cross your own chasm faster by standing on other people's crossings. It's the same reasoning behind why network effects are so powerful once they take hold.
Frequently Asked Questions
What is the chasm in Crossing the Chasm?
The chasm is the gap in the technology adoption life cycle between the early adopters (visionaries) and the early majority (pragmatists). Geoffrey Moore argued that these two groups buy for opposite reasons, so the enthusiasm that wins early adopters doesn't carry over to the mainstream. Many tech products get stuck in this gap: they win a passionate early following, then stall out because pragmatic mainstream buyers won't follow the same signals.
What is a beachhead market and why does it matter?
A beachhead market is one narrow, tightly defined segment of pragmatist customers who share an urgent problem, buy in similar ways, and talk to each other. Moore's advice is to dominate that single segment completely before expanding, using a D-Day analogy: concentrate all your force on one beach, own it, then push inland. It matters because pragmatists trust references from peers in their own niche, so winning a large share of one small market generates the word of mouth that carries you across the chasm.
What does "whole product" mean?
The whole product is everything a customer needs to fully solve their problem, not just the core feature you built. It includes integrations, support, training, documentation, partners, and standards. Visionaries will accept a bare-bones "generic product" and build the rest themselves, but pragmatists want the whole thing to already exist, because any missing piece reads to them as risk. Delivering the whole product, often through partnerships, is central to crossing the chasm.
Is Crossing the Chasm still relevant today?
Yes, arguably more than ever. Enterprise AI in the mid-2020s is a live example: consumer AI crossed almost instantly, but a 2025 MIT study found roughly 95% of corporate AI pilots delivered no measurable return. That split, easy early adoption but hard mainstream scaling, is exactly the chasm Moore described. The framework still explains why powerful products stall between early enthusiasts and pragmatic buyers, and how focus on a beachhead gets them across.
What's the difference between Crossing the Chasm and diffusion of innovations?
Everett Rogers' Diffusion of Innovations (1962) described the five adopter categories and the smooth bell curve of how new ideas spread. Moore borrowed that model but added a critical twist: for disruptive high-tech products, the curve isn't smooth. There's a chasm between early adopters and the early majority that companies must deliberately cross. Rogers himself pushed back on the chasm idea, but for founders and marketers, Moore's version has proven the more actionable of the two.
Conclusion: Cross on Purpose
The reason Crossing the Chasm has outlived its 1991 examples is that it names a failure mode people keep repeating. Early success feels like proof, so founders scale toward the mainstream as if it's more of the same crowd. It isn't. The mainstream is a different set of people with a different psychology, separated from your early fans by a gap that word of mouth won't jump on its own.
Moore's cure is discipline over ambition, at least for a while. Pick one beachhead. Build the whole product for it. Win enough of that niche to create references pragmatists believe. Then bowl into adjacent segments, and when the mainstream finally tips into a tornado, switch playbooks and grab the market. It's a sequence, and running the stages out of order is how good products die with great early reviews.
The through-line for anyone, founder or not, is judgment: the ability to read adoption honestly and see a chasm before you fall into it. That judgment is built, not born, and it's built by studying how real products crossed or failed. Start capturing those patterns with Glasp's web highlighter and YouTube Summary, turn your highlights into a searchable map of what actually works, and you'll be the one who crosses on purpose while everyone else is still celebrating their early adopters. For a related lens on why customers really "hire" a product, read Jobs to Be Done.