Crossing the Chasm, Explained by Its Author: Start Small, Win a Segment, Then Scale

Moore says his books are frameworks, models of the patterns to look for as something evolves, and that freedom is exactly why people misuse them. His example is a founder who declared that the Fortune 500 was their beachhead segment, which shows how easily companies look at the model from the inside out instead of from the customer's side.

The reasoning for staying narrow is about power. In an emerging category a startup needs enough leverage to steer its own course, and that comes first from a lighthouse customer and then from an ecosystem of partners. Ecosystems, Moore argues, form around market leaders and not around everyone else, so the only way to attract partners is to be a big fish in a small pond, with a realistic shot at 30 to 50 percent of a segment within two years.

He compares chasing several segments at once to running in several presidential primaries simultaneously: votes in Vermont do not count if you are campaigning in New Hampshire. The mechanism underneath is that pragmatic buyers purchase what they see their peers purchasing. If peers split across products A, B, C and D there is no leader and the category drifts sideways, but if three of four use the same product the rest follow.

Moore frames the shift in a single phrase. Before the chasm, the customer says "we believe what you believe," and after it, the customer says "we need what you have." Selling to visionaries means sharing a vision, while selling to pragmatists means showing you understand their problem.

The technology adoption lifecycle has four inflection points. The early market runs on a project model with a visionary executive sponsor who creates the budget, and it is worth landing a marquee name like Apple or Verizon because nobody notices a win with an unknown company. The second stage is the Crossing the Chasm playbook, organized around a problem that is being solved badly and getting worse, where the vendor over-commits to that problem and moves from projects to a repeatable solution.

From there the bowling alley extends to adjacent segments, either the same customer with a new use case or the same use case with a new customer base, which can take a company from tens to hundreds of millions of dollars. The third stage is the tornado, when a category goes horizontal and buyers say "we want what they have," so you need broad sales coverage and a standard product. The last is Main Street, where products are commoditized and innovation moves to services, as Uber did for taxis.

To explain why startups should resist taking any customer who will pay, Moore uses a bonfire analogy: running a match back and forth under a log never lights it, but holding it against kindling does. That is why adjacency matters in the bowling alley, because the next piece of kindling has to sit next to the fire and not in another room.

His favorite example is Documentum, a document database that started with pharmaceutical companies, where new drug approvals ran to hundreds of thousands of pages and every day lost cost a day of patent life. Chemical companies, then oil and gas with their leases, and then financial firms followed. Each was a new segment, but the use case stayed close enough to reuse.

Moore also notes that you normally cannot use the visionary's industry as your beachhead, because visionaries want to beat their own competitors, not help their industry. The formula he offers founders is a segment that is big enough to matter, small enough to lead, and a good fit with your crown jewels. Segments should share a geography, an industry and a profession, since those are the groups who actually talk to one another, and the target should offer room to reach about 100 million dollars within five years without being a billion-dollar pond.

Asked when a startup should attempt the chasm, Moore says revenue is the wrong measure. What matters is having a major account that has gone all in on the new technology and is willing to talk about it, which visionaries with healthy egos usually are. After that the play is inexpensive: no Super Bowl ads, just a message aimed at perhaps 200 people in one geography, industry and profession.

On the risk that big customers push a startup around, he says it depends on the persona of the executive sponsor. Nine in ten sponsors are company people who will send you to purchasing; the tenth is tired of the status quo and wants to leapfrog. Moore also advises starting with customer-led projects, aiming for cashflow break-even without venture capital, and raising venture money only when the technology is too expensive to self-fund or the market will catch fire too soon.

Lenny cites Moore's line that one round of funding should be enough to dominate a single use case in a single market within 18 to 24 months. Moore adds that because the problem is severe, you use value pricing and a guaranteed commitment to solve it, like choosing a heart surgeon over a coupon. Lenny's Figma and Coda story, where the team drove back to fix a Wi-Fi problem, illustrates the obsessive care he means.

Moore discovered the chasm while working at the marketing agency Regis McKenna, when companies that had won magazine covers vanished a couple of years later. Visionaries make buying decisions on their own and avoid doing what peers do, so these companies assumed visionaries could vouch for them to pragmatists. Pragmatists, however, see visionaries as people who think they are smarter, decide in ways pragmatists never would, and leave messes for others to clean up.

What pragmatists want is a peer reference, but no peer has gone first, which Moore calls the junior high dance problem. Lenny stresses that the reference customer must be a pragmatist who feels like "my person." Moore adds that people are visionaries about some purchases and conservatives about others, so the question is always what persona you hold for a particular decision.

On how much product is needed, a visionary only requires the magic ingredient, something roughly ten times better, even if the product is buggy. Pragmatists cannot live with a product that does not work, so you may need more project work until it is stable enough to productize. Only then should you find a market where it can be the dominant solution.

Buying software is a risk-reduction exercise, Moore says, echoing April Dunford's point that it can be harder than selling it. A pragmatist's first rule is that if it is not broken, do not fix it. Only a compelling reason to buy puts them under enough duress to move against their company's decision-making inertia, such as a leader who has already tried everyone internally and told the team the current answer is not good enough.

He offers examples including ransomware, a child struggling at school, healthcare, legal trouble and the post-pandemic question of what to do with office space. Lenny notes the common confusion between a compelling reason to buy and a compelling reason to sell, and Moore agrees that the pain is the point. Vendors who answer a lack of traction with a flashier demo or a new deck are working on the wrong side of the table.

In the early market and the tornado the conversation is about you, but in the bowling alley it never is. Moore's instruction is to shut the laptop and open with something like: we have been working with people in your industry, we understand there is a serious problem around this, do you have it? Customers either say yes or say "not exactly, our real problem is...", and that problem-domain knowledge is the gold you are after.

Moore explains how to recognize each stage. The early market needs a wizard technologist, a demo, and a vision of trapped value, since the world typically hands a company roughly ten percent of the value it releases. A billion-dollar company therefore needs about ten billion of trapped value, as Airbnb found in people's homes and Uber in their back seats.

For the bowling alley, you work down to a geography, industry, profession and use case and stay curious about why the problem is hard and what not solving it costs. The go-to-market follows from that, with lead generation built around symptoms, a business development rep confirming the problem, and a diagnostic call in which the customer talks and you take notes visibly with a pen. He suggests finding the right person, which may be Harry and not Mary.

On timing, Moore describes how venture capital really works: investors fund you to change the state of the company so that the next investor values it two to three times higher. Crossing the chasm means turning a cool possibility into a going concern, with loyal customers, partners who bring deals, and a worked-out CAC and LTV, probably at ten to twenty million dollars. The test is that you no longer need to raise, and if you do, you raise on your own timeline.

Moore says investors price risk. Angels take off the risk of whether you can create anything, chasm money takes off the risk of going out of business, and bowling alley money buys a journey from ten million toward a hundred million where the rule of 40 begins to matter. In the tornado the category itself is hot, which is why generative AI companies are valued so differently.

The difference between bowling alley and tornado is budget. Before the tornado, buyers either have no budget (early market) or budget earmarked for the old way to patch the problem (bowling alley), so you must create or redirect it. When a category goes horizontal, budgets appear everywhere at once, departments spend them, vendors are chosen, and ecosystems form around the winners. Moore points to Cisco, Intel, Oracle and Microsoft in the nineties and notes that a non-leader can retreat to being a chimp, a local gorilla, as Juniper did.

Qualifying on budget is critical in the tornado but a mistake in the early market or bowling alley, where it will never be met. A project model will not scale in the bowling alley because it is not repeatable, and a solution model over-invests in one thing for a visionary who wants to pull you off your roadmap. Moore says the response is not hard to see; the trouble is that people stay loyal to the playbook that made them successful.

Moore says AI is not one phase, because from a customer's view it exists in the early market, the bowling alley, the tornado and on Main Street. Microsoft Copilot is Main Street, a low-risk add-on to a suite people already use. Salesforce is the closest thing to a tornado use case, with sales, service and marketing copilots rolled out across a whole customer base on its own language model, bringing high productivity at modest risk.

For the bowling alley he cites Sal Khan and Khan Academy. After the pandemic a teacher may face five grade levels in one classroom, an impossible problem that tutoring tuned to each level could address. He also imagines specialist services that start with generative AI and keep a human in the loop, such as legal opinions or image design, which shows the technology can be absorbed at several places.

Discussing the seven deadly sins of crossing the chasm, Moore says discounting and freemium suit commoditized or no-risk purchases, but chasms are about risk-bearing decisions. Discounting does not reduce risk and might raise it if the vendor cuts support or fights over scope. Another sin is confusing the target customer: first find a small pool of trapped value, then identify who controls access to it and will sponsor your deal.

To validate a segment, Moore suggests a modest marketing campaign and a three-month bet on winning two more deals with the same pattern. He clarifies that the marquee customer is usually not in the beachhead market; it is the famous visionary win, like Han Solo's Kessel Run. Talk to the economic buyer, he says, because end users will complain happily but cannot sponsor anything.

Pragmatists in the chasm rarely say no, he warns; they just never say yes, and encourage you to come back. His positioning formula is the same every time: you are the technology leader committed to solving this problem, you respect the incumbent vendor who lacks the technology, and you respect technology peers who lack the domain expertise. Nobody in your quadrant will handle the problem as well as you.

For bottom-up companies like Atlassian or Yammer, Moore says the game differs because you attract the end user before the economic buyer. You need telemetry to see what people do and a way to find a compelling reason to buy inside their environment. But he says you cannot cross the chasm with product-led growth, because pragmatists need to learn more, in the same way you cannot beat Covid by leaving vaccines in public places.

Where product-led growth shines is the land and expand phases, because a hot product prompts users to get more involved, which is a Main Street move that works when the next purchase carries little risk. Eventually such companies build sales teams to win enterprise deals, but Moore warns against hiring a typical enterprise salesperson while crossing the chasm. He wants someone closer to a sales engineer, diagnostic and committed to the integrity of the problem and solution.

Asked about his exchange with Martin Casado of Andreessen Horowitz, Moore says Casado's main point was that venture investors no longer work at this level of granularity, with so much capital and so much existing software. Casado also argued that life never gets easy after the chasm, and Moore agrees: the problems just change, with competition, funding, and shifts from data center to cloud to Kubernetes to edge AI. There will be a new headache every week.

Moore says the model is optimized for B2B markets with federated, high-risk buying decisions, which accounted for roughly 95 percent of tech in the twentieth century. Consumer computing then arrived from nowhere, and in a world where B2C can be the core of innovation, he does not think the chasm model was designed for that problem. For B2B, he says, it remains the most reliable playbook.

In closing, Moore encourages entrepreneurs to make a great living but then pursue impact, since someone gifted enough to start a software company is a scarce resource. He also describes his book The Infinite Staircase, which traces 11 steps from physics to human conversation and argues that ethics can come from below, from how mammals nurture their young, and not only from above. He is on LinkedIn, where he also keeps a blog.