Crossing the Chasm is a Meme, but not a Principle of Economic Science

“Everything should be made as simple as possible. But not simpler.” — Albert Einstein

The ”crossing the chasm” hypothesis is an elegant idea, it’s well defined and certainly has a lot of relevance in describing how technology is adopted by markets. However, is it too simple? Martin Casado asked that question in his article “ The Unending Chasm, and How to Survive it “. Martin’s experience matches substantially with mine. That is, while “Crossing the Chasm” is useful, it often doesn’t provide useful guidance to someone entering a new market.

In this post, I describe my experiences as an executive in two companies bringing technology to new markets. The underlying goal is to further show that a general model like “Crossing the Chasm” isn’t as directly useful as one might wish.

Too much market pull can kill a company

Let’s start with the first example. Some time ago, I joined a startup which invented a new product for a new, high growing market, but with tremendous pressure from a pricing perspective. Everybody wanted to buy the product before it was ready.

So we were immediately facing a “pull” based market. The customers tested the prototype, loved it, and wanted to buy it if the price and quality were right. We were selling the product and sales were in single digits millions. We had the fortune of not having to “push”. We had the use case and we just needed the product to fit the price range.

There was a large competitor in the market who was not bothered by this new kid on the block, not because the market was too small, but because the product was difficult to build. Once we built the right product, it started selling like hot cakes, sales quickly reaching hundreds of millions, and then half a billion.

We crossed the chasm but we were on top of a cliff! The revenue was growing fast to the hundreds of millions, but losses were growing even faster. Customers were “pulling”, but they were pulling us into a crevasse. The company had to slow down, decrease production, improve the gross margin and now faced yet another chasm to cross. On every new international market, with every new product added to the portfolio we needed to cross a new chasm, because every international market had its own incumbent, which adapted to new technology. And all this was happening while customers were continuing to pull. So in summary:

1. We had market pull before building a product

1. At first we did not have any chasm in front of us
2. The market pull was like a black hole forcing us to manufacture more and more, but we were losing money (having low gross margins) and we were in trouble.
3. We had to go back to “early adopters” mode
4. In these different conditions, indeed we had a chasm in front of us due to harsh competition and the diminished pull
5. We limited sales, but increased the gross margin
6. The company is selling around $500M/year but has not crossed the last chasm; it is doing adequately, slowly climbing to $1B market cap.

Aggressive tactical “push” can help crossing the chasm and generate pull, but it may create a product that is hard to evolve.

Second example: This was a company in a market space with very large incumbents. Initially it was a startup; for some time it struggled to find a use case with a simple product. The company talked to customers and tried different features. The product kept changing and the company tried to find a use case and insertion strategy. As a result the team neglected the architecture and transformed the product into a “Christmas tree” of features, which was working reasonably well.

Then suddenly, the company found the use case. With that, it entered a push mode and crossed the chasm with customers pulling the product. But it was a niche market and after several years of growth (going to $2B market cap) sales started to saturate at less than $200M. However, the company was successful and employees happy.

Then the market changed. An aggressive newcomer had a well architected product with different features and was overlapping part of our market. We had to re-adapt the product but it was difficult to change. Our market cap collapsed while our revenue remained constant with +70% gross margin, making good money. The company had continuous pull from customers, however the pull was stagnant. We were successful in what we set out to do, but the outside world’s perception that another superior competitor may get our market share (which was not true) was not helping us. Just imagine a two lane road: a slow lane with only one car moving with constant and medium speed and another lane with several cars passing each other at high speed. Both lanes were the road to success. If we tried to adapt the case to Moore’s crossing the chasm paradigm, from product perspective the company was always inside the chasm and from sales perspective was outside the chasm, in pull mode. We work hard for several year to create a new, well architected, world class product line and succeeded in our efforts. But we had to cross again another chasm. In summary:

1. We crossed the chasm with a working product, but with broken architecture;
2. The pull was continuous but stagnant;
3. We reached market saturation and entered into complacency;
4. An aggressive competitor jumped into an adjacent market space
5. We had to go back into chasm even if customer pull was there and we were making great money.

In the open source world incumbents are nimble and innovative

The idea that large incumbents are slow, not innovative, and trapped inside their own projects, currently, is incorrect. Open source started with the initiative of “lone wolves”, researchers or academics who created software useful for a limited community (Unix, Linux, etc). Today, open source is taking the world by storm. However, this is not because startups are generating tremendous open source value; in fact, venture capitalists do not invest in brand new open source projects, but prefer to jump on projects that have already years of open source success and large penetration in the developers’ community. Many of the most relevant open source projects have been sponsored or created inside large enterprises: Kubernetes, Hadoop, Kafka, etc. There is, actually, a strategy of large companies encouraging their engineers to contribute to or create open source projects. On careful analysis, large corporations are generating many billions in revenue by building proprietary products on top of open source software: AWS, Facebook, Twitter, Google, Uber, Apple, Microsoft, etc. From a financial perspective, open source is probably not very fair for individual contributors and early startups. It is becoming more and more difficult to innovate in open source as a startup. Incumbents are interested in creating adjacent market spaces of small initial TAM of $50M- $100M, if those markets can significantly grow and there are many of them. The myth that large companies are only interested in more than $1B revenue potentials for their products is not true. Good examples are cloud computing companies.

The business world is not linear, but a multi-dimensional and dynamic system

Underlying business topology has several dimensions such as geographical space, time scale, competitors, different classes of customers, and vertical markets. This means that, if the “crossing the chasm” meme is correct, then the chasm is multi-dimensional, or N–dimensional.

On top of this N-dimensional aspect, the business topology is also variable: market conditions change rapidly, competitors are diverse, unpredictable, and fast-moving, and incumbents are moving also with fast pace. This means that customer “pull” may come and go in rapid bursts, sometimes at large scale, and the company has to continuously be in “push” mode. Mathematically speaking, a “crossing the chasm” idea or concept could, probably, be modeled as an optimization problem with optimization goals (as specified by Moore) being a vector of with several dimensions:

MooreVector = (use cases, right products, word of mouth)

To be solvable, the optimization problem requires the definition of the vectors of constraints and demands (all dynamic). Detailed research is needed for a correct classification of the problem, but intuitively “crossing the chasm” seems to be a NP problem (non-deterministic polynomial time).

Any intuitive reflection without the backing of a mathematical proof remains just a logical reflection, shiny, in some cases, but with limited practical applicability. Example: it is well known in the history of science that scientists used mental experiments to explain mathematical models that started by being intuitive, but ultimately, well thought mathematical models were discovered or established to justify them. Aristotle, Galileo, and Einstein are well known personalities who used this methodology of combining logical experiments with mathematical forms of proof.

To be successful, businesses have to be always prepared for battle

“In preparing for battle I have always found that plans are useless, but planning is indispensable.” –Dwight Eisenhower

Companies need to continuously be in push mode even if there is a strong pull from customers. A strong customer pull can drive them into complacency and, ultimately, to their failure.

Some companies can be successful (success having several definitions) without ever getting into “customer pull” mode, and remaining in “eternal push” mode, just because sales cycles have a long time-scale or move with time.

When companies have products that reached market saturation, they need to reinvent themselves, find other use cases, invent new products, market them with “word of mouth”, and thus be in a “push-pull” mode forever in order to survive.

If there is no science behind an economic idea, we should be skeptical

There are four fundamental sciences: mathematics, physics, chemistry, and biology. The evolution of the Universe and Humankind is based on the laws of these four pillars and other sciences derived from them. Economic sciences are derived from social sciences and mathematics. The “crossing the chasm” observation can be framed as a part of growth models. Even if “the chasm” is represented on a Gaussian bell-curve, there is no mathematical definition or justification of it. Moreover, statistics allows economics to make forecasts and determine the probability of an occurrence. What is the probability of “crossing the chasm”? What is the probability of “the chasm’s” occurrence?

The duality theory-experiment is essential. It is true that there are examples in the science of theorems which were accepted as true without a proof. The Last Theorem of Fermat (ultimately a proof was found) and Riemann’s conjecture are wonderful examples of this. There are also examples of principles which had a proof but did not have experimental verification. Theory of relativity and quantum entanglement are other examples. Until Sir Arthur Eddington proved that the gravitational field curves light, the theory of relativity was not widely accepted. Darwin’s Law of Evolution is primordially an experimental law based on observation and statistics but also with mathematical representation.

We have to be skeptical of models that do not have a profound scientific definition and experimental or theoretical proof. We need to ask questions and try to find correct answers, which are verifiable and proved theoretically and/or experimentally with a very high probability of occurrence.

Geoffrey Moore is correct when talking about “Crossing the Chasm” as a meme. Many in the industry have transformed it into a universal principle, into a myth, neglecting that there is no solid scientific definition or justification of it. That is simply wrong.

What Martin Casado did with his article “ The Unending Chasm, and How to Survive it “ is to map it to his experience. He observed experiments, started to doubt, and presented counter-examples. This methodology has been a catalyst for many discoveries in the history of science and was described very simply by René Descartes:

“Dubito, ergo cogit”.